diff --git a/.claude-plugin/marketplace.json b/.claude-plugin/marketplace.json index 568f24331..a8a6b362a 100644 --- a/.claude-plugin/marketplace.json +++ b/.claude-plugin/marketplace.json @@ -5,14 +5,14 @@ }, "metadata": { "description": "Headroom marketplace for Claude Code and GitHub Copilot CLI plugins.", - "version": "0.20.2" + "version": "0.20.8" }, "plugins": [ { "name": "headroom", "source": "./plugins/headroom-agent-hooks", "description": "Headroom startup hooks for Claude Code and GitHub Copilot CLI.", - "version": "0.20.2", + "version": "0.20.8", "author": { "name": "Headroom Contributors", "url": "https://github.com/chopratejas/headroom" diff --git a/.github/plugin/marketplace.json b/.github/plugin/marketplace.json index 568f24331..a8a6b362a 100644 --- a/.github/plugin/marketplace.json +++ b/.github/plugin/marketplace.json @@ -5,14 +5,14 @@ }, "metadata": { "description": "Headroom marketplace for Claude Code and GitHub Copilot CLI plugins.", - "version": "0.20.2" + "version": "0.20.8" }, "plugins": [ { "name": "headroom", "source": "./plugins/headroom-agent-hooks", "description": "Headroom startup hooks for Claude Code and GitHub Copilot CLI.", - "version": "0.20.2", + "version": "0.20.8", "author": { "name": "Headroom Contributors", "url": "https://github.com/chopratejas/headroom" diff --git a/.gitignore b/.gitignore index dce3a18a1..42a73b1c7 100644 --- a/.gitignore +++ b/.gitignore @@ -22,6 +22,7 @@ scripts/* !scripts/build_rust_extension.sh !scripts/install-git-hooks.sh !scripts/smoke_issue_327.py +!scripts/refresh_model_limits.sh # Rust / Cargo build artifacts /target/ diff --git a/crates/headroom-proxy/data/model_prices_and_context_window.json b/crates/headroom-proxy/data/model_prices_and_context_window.json new file mode 100644 index 000000000..bbe13442d --- /dev/null +++ b/crates/headroom-proxy/data/model_prices_and_context_window.json @@ -0,0 +1,39767 @@ +{ + "sample_spec": { + "code_interpreter_cost_per_session": 0.0, + "computer_use_input_cost_per_1k_tokens": 0.0, + "computer_use_output_cost_per_1k_tokens": 0.0, + "deprecation_date": "date when the model becomes deprecated in the format YYYY-MM-DD", + "file_search_cost_per_1k_calls": 0.0, + "file_search_cost_per_gb_per_day": 0.0, + "input_cost_per_audio_token": 0.0, + "input_cost_per_token": 0.0, + "litellm_provider": "one of https://docs.litellm.ai/docs/providers", + "max_input_tokens": "max input tokens, if the provider specifies it. if not default to max_tokens", + "max_output_tokens": "max output tokens, if the provider specifies it. if not default to max_tokens", + "max_tokens": "LEGACY parameter. set to max_output_tokens if provider specifies it. IF not set to max_input_tokens, if provider specifies it.", + "mode": "one of: chat, embedding, completion, image_generation, audio_transcription, audio_speech, image_generation, moderation, rerank, search", + "output_cost_per_reasoning_token": 0.0, + "output_cost_per_token": 0.0, + "search_context_cost_per_query": { + "search_context_size_high": 0.0, + "search_context_size_low": 0.0, + "search_context_size_medium": 0.0 + }, + "supported_regions": [ + "global", + "us-west-2", + "eu-west-1", + "ap-southeast-1", + "ap-northeast-1" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true, + "vector_store_cost_per_gb_per_day": 0.0 + }, + "1024-x-1024/50-steps/bedrock/amazon.nova-canvas-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 2600, + "mode": "image_generation", + "output_cost_per_image": 0.06 + }, + "1024-x-1024/50-steps/stability.stable-diffusion-xl-v1": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.04 + }, + "1024-x-1024/dall-e-2": { + "input_cost_per_pixel": 1.9e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "1024-x-1024/max-steps/stability.stable-diffusion-xl-v1": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.08 + }, + "256-x-256/dall-e-2": { + "input_cost_per_pixel": 2.4414e-07, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "512-x-512/50-steps/stability.stable-diffusion-xl-v0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.018 + }, + "512-x-512/dall-e-2": { + "input_cost_per_pixel": 6.86e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "512-x-512/max-steps/stability.stable-diffusion-xl-v0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.036 + }, + "ai21.j2-mid-v1": { + "input_cost_per_token": 1.25e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 8191, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 1.25e-05 + }, + "ai21.j2-ultra-v1": { + "input_cost_per_token": 1.88e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 8191, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 1.88e-05 + }, + "ai21.jamba-1-5-large-v1:0": { + "input_cost_per_token": 2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06 + }, + "ai21.jamba-1-5-mini-v1:0": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07 + }, + "ai21.jamba-instruct-v1:0": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 70000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7e-07, + "supports_system_messages": true + }, + "aiml/dall-e-2": { + "litellm_provider": "aiml", + "metadata": { + "notes": "DALL-E 2 via AI/ML API - Reliable text-to-image generation" + }, + "mode": "image_generation", + "output_cost_per_image": 0.026, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/dall-e-3": { + "litellm_provider": "aiml", + "metadata": { + "notes": "DALL-E 3 via AI/ML API - High-quality text-to-image generation" + }, + "mode": "image_generation", + "output_cost_per_image": 0.052, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux-pro": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Flux Dev - Development version optimized for experimentation" + }, + "mode": "image_generation", + "output_cost_per_image": 0.065, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux-pro/v1.1": { + "litellm_provider": "aiml", + "mode": "image_generation", + "output_cost_per_image": 0.052, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux-pro/v1.1-ultra": { + "litellm_provider": "aiml", + "mode": "image_generation", + "output_cost_per_image": 0.063, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux-realism": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Flux Pro - Professional-grade image generation model" + }, + "mode": "image_generation", + "output_cost_per_image": 0.046, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux/dev": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Flux Dev - Development version optimized for experimentation" + }, + "mode": "image_generation", + "output_cost_per_image": 0.033, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux/kontext-max/text-to-image": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Flux Pro v1.1 - Enhanced version with improved capabilities and 6x faster inference speed" + }, + "mode": "image_generation", + "output_cost_per_image": 0.104, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux/kontext-pro/text-to-image": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Flux Pro v1.1 - Enhanced version with improved capabilities and 6x faster inference speed" + }, + "mode": "image_generation", + "output_cost_per_image": 0.052, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/flux/schnell": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Flux Schnell - Fast generation model optimized for speed" + }, + "mode": "image_generation", + "output_cost_per_image": 0.004, + "source": "https://docs.aimlapi.com/", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/google/imagen-4.0-ultra-generate-001": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Imagen 4.0 Ultra Generate API - Photorealistic image generation with precise text rendering" + }, + "mode": "image_generation", + "output_cost_per_image": 0.078, + "source": "https://docs.aimlapi.com/api-references/image-models/google/imagen-4-ultra-generate", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "aiml/google/nano-banana-pro": { + "litellm_provider": "aiml", + "metadata": { + "notes": "Gemini 3 Pro Image (Nano Banana Pro) - Advanced text-to-image generation with reasoning and 4K resolution support" + }, + "mode": "image_generation", + "output_cost_per_image": 0.195, + "source": "https://docs.aimlapi.com/api-references/image-models/google/gemini-3-pro-image-preview", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "amazon.nova-canvas-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 2600, + "mode": "image_generation", + "output_cost_per_image": 0.06, + "supports_nova_canvas_image_edit": true + }, + "us.amazon.nova-canvas-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 2600, + "mode": "image_generation", + "output_cost_per_image": 0.06, + "supports_nova_canvas_image_edit": true + }, + "us.writer.palmyra-x4-v1:0": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_pdf_input": true + }, + "us.writer.palmyra-x5-v1:0": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_pdf_input": true + }, + "writer.palmyra-x4-v1:0": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_pdf_input": true + }, + "writer.palmyra-x5-v1:0": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_pdf_input": true + }, + "amazon.nova-lite-v1:0": { + "input_cost_per_token": 6e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 2.4e-07, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "amazon.nova-2-lite-v1:0": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "amazon.nova-2-pro-preview-20251202-v1:0": { + "cache_read_input_token_cost": 5.46875e-07, + "input_cost_per_token": 2.1875e-06, + "input_cost_per_image_token": 2.1875e-06, + "input_cost_per_audio_token": 2.1875e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.75e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "apac.amazon.nova-2-lite-v1:0": { + "cache_read_input_token_cost": 8.25e-08, + "input_cost_per_token": 3.3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "apac.amazon.nova-2-pro-preview-20251202-v1:0": { + "cache_read_input_token_cost": 5.46875e-07, + "input_cost_per_token": 2.1875e-06, + "input_cost_per_image_token": 2.1875e-06, + "input_cost_per_audio_token": 2.1875e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.75e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "eu.amazon.nova-2-lite-v1:0": { + "cache_read_input_token_cost": 8.25e-08, + "input_cost_per_token": 3.3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "eu.amazon.nova-2-pro-preview-20251202-v1:0": { + "cache_read_input_token_cost": 5.46875e-07, + "input_cost_per_token": 2.1875e-06, + "input_cost_per_image_token": 2.1875e-06, + "input_cost_per_audio_token": 2.1875e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.75e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "us.amazon.nova-2-lite-v1:0": { + "cache_read_input_token_cost": 8.25e-08, + "input_cost_per_token": 3.3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "us.amazon.nova-2-pro-preview-20251202-v1:0": { + "cache_read_input_token_cost": 5.46875e-07, + "input_cost_per_token": 2.1875e-06, + "input_cost_per_image_token": 2.1875e-06, + "input_cost_per_audio_token": 2.1875e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.75e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "amazon.nova-2-multimodal-embeddings-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 8172, + "max_tokens": 8172, + "mode": "embedding", + "input_cost_per_token": 1.35e-07, + "input_cost_per_image": 6e-05, + "input_cost_per_video_per_second": 0.0007, + "input_cost_per_audio_per_second": 0.00014, + "output_cost_per_token": 0.0, + "output_vector_size": 3072, + "source": "https://us-east-1.console.aws.amazon.com/bedrock/home?region=us-east-1#/model-catalog/serverless/amazon.nova-2-multimodal-embeddings-v1:0", + "supports_embedding_image_input": true, + "supports_image_input": true, + "supports_video_input": true, + "supports_audio_input": true + }, + "amazon.nova-micro-v1:0": { + "input_cost_per_token": 3.5e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.4e-07, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true + }, + "amazon.nova-pro-v1:0": { + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "amazon.rerank-v1:0": { + "input_cost_per_query": 0.001, + "input_cost_per_token": 0.0, + "litellm_provider": "bedrock", + "max_document_chunks_per_query": 100, + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_query_tokens": 32000, + "max_tokens": 32000, + "max_tokens_per_document_chunk": 512, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "amazon.titan-embed-image-v1": { + "input_cost_per_image": 6e-05, + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128, + "max_tokens": 128, + "metadata": { + "notes": "'supports_image_input' is a deprecated field. Use 'supports_embedding_image_input' instead." + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://us-east-1.console.aws.amazon.com/bedrock/home?region=us-east-1#/providers?model=amazon.titan-image-generator-v1", + "supports_embedding_image_input": true, + "supports_image_input": true + }, + "amazon.titan-embed-text-v1": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536 + }, + "amazon.titan-embed-text-v2:0": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024 + }, + "amazon.titan-image-generator-v1": { + "input_cost_per_image": 0.0, + "output_cost_per_image": 0.008, + "output_cost_per_image_premium_image": 0.01, + "output_cost_per_image_above_512_and_512_pixels": 0.01, + "output_cost_per_image_above_512_and_512_pixels_and_premium_image": 0.012, + "litellm_provider": "bedrock", + "mode": "image_generation" + }, + "amazon.titan-image-generator-v2": { + "input_cost_per_image": 0.0, + "output_cost_per_image": 0.008, + "output_cost_per_image_premium_image": 0.01, + "output_cost_per_image_above_1024_and_1024_pixels": 0.01, + "output_cost_per_image_above_1024_and_1024_pixels_and_premium_image": 0.012, + "litellm_provider": "bedrock", + "mode": "image_generation" + }, + "amazon.titan-image-generator-v2:0": { + "input_cost_per_image": 0.0, + "output_cost_per_image": 0.008, + "output_cost_per_image_premium_image": 0.01, + "output_cost_per_image_above_1024_and_1024_pixels": 0.01, + "output_cost_per_image_above_1024_and_1024_pixels_and_premium_image": 0.012, + "litellm_provider": "bedrock", + "mode": "image_generation" + }, + "twelvelabs.marengo-embed-2-7-v1:0": { + "input_cost_per_token": 7e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "supports_embedding_image_input": true, + "supports_image_input": true + }, + "us.twelvelabs.marengo-embed-2-7-v1:0": { + "input_cost_per_token": 7e-05, + "input_cost_per_video_per_second": 0.0007, + "input_cost_per_audio_per_second": 0.00014, + "input_cost_per_image": 0.0001, + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "supports_embedding_image_input": true, + "supports_image_input": true + }, + "eu.twelvelabs.marengo-embed-2-7-v1:0": { + "input_cost_per_token": 7e-05, + "input_cost_per_video_per_second": 0.0007, + "input_cost_per_audio_per_second": 0.00014, + "input_cost_per_image": 0.0001, + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "supports_embedding_image_input": true, + "supports_image_input": true + }, + "twelvelabs.pegasus-1-2-v1:0": { + "input_cost_per_video_per_second": 0.00049, + "output_cost_per_token": 7.5e-06, + "litellm_provider": "bedrock", + "mode": "chat", + "supports_video_input": true + }, + "us.twelvelabs.pegasus-1-2-v1:0": { + "input_cost_per_video_per_second": 0.00049, + "output_cost_per_token": 7.5e-06, + "litellm_provider": "bedrock", + "mode": "chat", + "supports_video_input": true + }, + "eu.twelvelabs.pegasus-1-2-v1:0": { + "input_cost_per_video_per_second": 0.00049, + "output_cost_per_token": 7.5e-06, + "litellm_provider": "bedrock", + "mode": "chat", + "supports_video_input": true + }, + "amazon.titan-text-express-v1": { + "input_cost_per_token": 1.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1.7e-06 + }, + "amazon.titan-text-lite-v1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 4e-07 + }, + "amazon.titan-text-premier-v1:0": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.5e-06 + }, + "anthropic.claude-3-5-haiku-20241022-v1:0": { + "cache_creation_input_token_cost": 1e-06, + "cache_read_input_token_cost": 8e-08, + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_creation_input_token_cost_above_1hr": 2e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "anthropic.claude-haiku-4-5@20251001": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_creation_input_token_cost_above_1hr": 2e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_streaming": true, + "supports_native_structured_output": true + }, + "anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 3e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "cache_creation_input_token_cost_above_1hr": 7.5e-06, + "cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.5e-05, + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07 + }, + "anthropic.claude-3-5-sonnet-20241022-v2:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 3e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "cache_creation_input_token_cost_above_1hr": 7.5e-06, + "cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.5e-05 + }, + "anthropic.claude-3-7-sonnet-20240620-v1:0": { + "cache_creation_input_token_cost": 4.5e-06, + "cache_read_input_token_cost": 3.6e-07, + "input_cost_per_token": 3.6e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.8e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "anthropic.claude-3-7-sonnet-20250219-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "anthropic.claude-3-haiku-20240307-v1:0": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 2.5e-08, + "cache_creation_input_token_cost": 3.125e-07 + }, + "anthropic.claude-3-opus-20240229-v1:0": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 1.5e-06, + "cache_creation_input_token_cost": 1.875e-05 + }, + "anthropic.claude-3-sonnet-20240229-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "anthropic.claude-instant-v1": { + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-06, + "supports_tool_choice": true + }, + "anthropic.claude-opus-4-1-20250805-v1:0": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "anthropic.claude-opus-4-20250514-v1:0": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "anthropic.claude-opus-4-5-20251101-v1:0": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_native_structured_output": true + }, + "anthropic.claude-opus-4-6-v1": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "global.anthropic.claude-opus-4-6-v1": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "us.anthropic.claude-opus-4-6-v1": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_1hr": 1.1e-05, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "eu.anthropic.claude-opus-4-6-v1": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "au.anthropic.claude-opus-4-6-v1": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "anthropic.claude-opus-4-7": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "anthropic.claude-mythos-preview": { + "input_cost_per_token": 0, + "output_cost_per_token": 0, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_vision": true, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "global.anthropic.claude-opus-4-7": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "us.anthropic.claude-opus-4-7": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_1hr": 1.1e-05, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "eu.anthropic.claude-opus-4-7": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "au.anthropic.claude-opus-4-7": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "anthropic.claude-sonnet-4-6": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_minimal_reasoning_effort": true + }, + "global.anthropic.claude-sonnet-4-6": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_minimal_reasoning_effort": true + }, + "us.anthropic.claude-sonnet-4-6": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_creation_input_token_cost_above_1hr": 6.6e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_minimal_reasoning_effort": true + }, + "eu.anthropic.claude-sonnet-4-6": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_minimal_reasoning_effort": true + }, + "au.anthropic.claude-sonnet-4-6": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_minimal_reasoning_effort": true + }, + "anthropic.claude-sonnet-4-20250514-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.2e-05, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_native_structured_output": true + }, + "anthropic.claude-v1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05 + }, + "anthropic.claude-v2:1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "anyscale/HuggingFaceH4/zephyr-7b-beta": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.5e-07 + }, + "anyscale/codellama/CodeLlama-34b-Instruct-hf": { + "input_cost_per_token": 1e-06, + "litellm_provider": "anyscale", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-06 + }, + "anyscale/codellama/CodeLlama-70b-Instruct-hf": { + "input_cost_per_token": 1e-06, + "litellm_provider": "anyscale", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-06, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/codellama-CodeLlama-70b-Instruct-hf" + }, + "anyscale/google/gemma-7b-it": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/google-gemma-7b-it" + }, + "anyscale/meta-llama/Llama-2-13b-chat-hf": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.5e-07 + }, + "anyscale/meta-llama/Llama-2-70b-chat-hf": { + "input_cost_per_token": 1e-06, + "litellm_provider": "anyscale", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-06 + }, + "anyscale/meta-llama/Llama-2-7b-chat-hf": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-07 + }, + "anyscale/meta-llama/Meta-Llama-3-70B-Instruct": { + "input_cost_per_token": 1e-06, + "litellm_provider": "anyscale", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-06, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/meta-llama-Meta-Llama-3-70B-Instruct" + }, + "anyscale/meta-llama/Meta-Llama-3-8B-Instruct": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/meta-llama-Meta-Llama-3-8B-Instruct" + }, + "anyscale/mistralai/Mistral-7B-Instruct-v0.1": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/mistralai-Mistral-7B-Instruct-v0.1", + "supports_function_calling": true + }, + "anyscale/mistralai/Mixtral-8x22B-Instruct-v0.1": { + "input_cost_per_token": 9e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/mistralai-Mixtral-8x22B-Instruct-v0.1", + "supports_function_calling": true + }, + "anyscale/mistralai/Mixtral-8x7B-Instruct-v0.1": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "anyscale", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://docs.anyscale.com/preview/endpoints/text-generation/supported-models/mistralai-Mixtral-8x7B-Instruct-v0.1", + "supports_function_calling": true + }, + "apac.amazon.nova-lite-v1:0": { + "input_cost_per_token": 6.3e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 2.52e-07, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "apac.amazon.nova-micro-v1:0": { + "input_cost_per_token": 3.7e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.48e-07, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true + }, + "apac.amazon.nova-pro-v1:0": { + "input_cost_per_token": 8.4e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.36e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "apac.anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "apac.anthropic.claude-3-5-sonnet-20241022-v2:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "apac.anthropic.claude-3-haiku-20240307-v1:0": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 2.5e-08, + "cache_creation_input_token_cost": 3.125e-07 + }, + "apac.anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.375e-06, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5.5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "apac.anthropic.claude-3-sonnet-20240229-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "apac.anthropic.claude-sonnet-4-20250514-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "assemblyai/best": { + "input_cost_per_second": 3.333e-05, + "litellm_provider": "assemblyai", + "mode": "audio_transcription", + "output_cost_per_second": 0.0 + }, + "assemblyai/nano": { + "input_cost_per_second": 0.00010278, + "litellm_provider": "assemblyai", + "mode": "audio_transcription", + "output_cost_per_second": 0.0 + }, + "au.anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "input_cost_per_token_above_200k_tokens": 6.6e-06, + "output_cost_per_token_above_200k_tokens": 2.475e-05, + "cache_creation_input_token_cost_above_200k_tokens": 8.25e-06, + "cache_read_input_token_cost_above_200k_tokens": 6.6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "azure/ada": { + "input_cost_per_token": 1e-07, + "litellm_provider": "azure", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "azure/codex-mini": { + "cache_read_input_token_cost": 3.75e-07, + "input_cost_per_token": 1.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 6e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/command-r-plus": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true + }, + "azure_ai/claude-haiku-4-5": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_creation_input_token_cost_above_1hr": 2e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/claude-opus-4-5": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/claude-opus-4-6": { + "input_cost_per_token": 5e-06, + "output_cost_per_token": 2.5e-05, + "litellm_provider": "azure_ai", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "azure_ai/claude-opus-4-7": { + "input_cost_per_token": 5e-06, + "output_cost_per_token": 2.5e-05, + "litellm_provider": "azure_ai", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 159, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "azure_ai/claude-opus-4-1": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_creation_input_token_cost_above_1hr": 3e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "azure_ai", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/claude-sonnet-4-5": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/claude-sonnet-4-6": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_minimal_reasoning_effort": true + }, + "azure/computer-use-preview": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/container": { + "code_interpreter_cost_per_session": 0.03, + "litellm_provider": "azure", + "mode": "chat" + }, + "azure_ai/gpt-oss-120b": { + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "source": "https://azure.microsoft.com/en-us/pricing/details/cognitive-services/openai-service/", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "azure_ai/model_router": { + "input_cost_per_token": 1.4e-07, + "output_cost_per_token": 0, + "litellm_provider": "azure_ai", + "mode": "chat", + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-services/", + "comment": "Flat cost of $0.14 per M input tokens for Azure AI Foundry Model Router infrastructure. Use pattern: azure_ai/model_router/ where deployment-name is your Azure deployment (e.g., azure-model-router)" + }, + "azure/eu/gpt-4o-2024-08-06": { + "deprecation_date": "2026-02-27", + "cache_read_input_token_cost": 1.375e-06, + "input_cost_per_token": 2.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-4o-2024-11-20": { + "deprecation_date": "2026-03-01", + "cache_creation_input_token_cost": 1.38e-06, + "input_cost_per_token": 2.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-4o-mini-2024-07-18": { + "cache_read_input_token_cost": 8.3e-08, + "input_cost_per_token": 1.65e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6.6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-4o-mini-realtime-preview-2024-12-17": { + "cache_creation_input_audio_token_cost": 3.3e-07, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_audio_token": 1.1e-05, + "input_cost_per_token": 6.6e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2.2e-05, + "output_cost_per_token": 2.64e-06, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/eu/gpt-4o-realtime-preview-2024-10-01": { + "cache_creation_input_audio_token_cost": 2.2e-05, + "cache_read_input_token_cost": 2.75e-06, + "input_cost_per_audio_token": 0.00011, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 0.00022, + "output_cost_per_token": 2.2e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/eu/gpt-4o-realtime-preview-2024-12-17": { + "cache_read_input_audio_token_cost": 2.5e-06, + "cache_read_input_token_cost": 2.75e-06, + "input_cost_per_audio_token": 4.4e-05, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 2.2e-05, + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/eu/gpt-5-2025-08-07": { + "cache_read_input_token_cost": 1.375e-07, + "input_cost_per_token": 1.375e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-5-mini-2025-08-07": { + "cache_read_input_token_cost": 2.75e-08, + "input_cost_per_token": 2.75e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-5.1": { + "cache_read_input_token_cost": 1.4e-07, + "input_cost_per_token": 1.38e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/eu/gpt-5.1-chat": { + "cache_read_input_token_cost": 1.4e-07, + "input_cost_per_token": 1.38e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/eu/gpt-5.1-codex": { + "cache_read_input_token_cost": 1.4e-07, + "input_cost_per_token": 1.38e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-5.1-codex-mini": { + "cache_read_input_token_cost": 2.8e-08, + "input_cost_per_token": 2.75e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 2.2e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/gpt-5-nano-2025-08-07": { + "cache_read_input_token_cost": 5.5e-09, + "input_cost_per_token": 5.5e-08, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4.4e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/o1-2024-12-17": { + "cache_read_input_token_cost": 8.25e-06, + "input_cost_per_token": 1.65e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6.6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/eu/o1-mini-2024-09-12": { + "cache_read_input_token_cost": 6.05e-07, + "input_cost_per_token": 1.21e-06, + "input_cost_per_token_batches": 6.05e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.84e-06, + "output_cost_per_token_batches": 2.42e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_vision": false + }, + "azure/eu/o1-preview-2024-09-12": { + "cache_read_input_token_cost": 8.25e-06, + "input_cost_per_token": 1.65e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6.6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_vision": false + }, + "azure/eu/o3-mini-2025-01-31": { + "cache_read_input_token_cost": 6.05e-07, + "input_cost_per_token": 1.21e-06, + "input_cost_per_token_batches": 6.05e-07, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.84e-06, + "output_cost_per_token_batches": 2.42e-06, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/global-standard/gpt-4o-2024-08-06": { + "cache_read_input_token_cost": 1.25e-06, + "deprecation_date": "2026-02-27", + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/global-standard/gpt-4o-2024-11-20": { + "cache_read_input_token_cost": 1.25e-06, + "deprecation_date": "2026-03-01", + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/global-standard/gpt-4o-mini": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/global/gpt-4o-2024-08-06": { + "deprecation_date": "2026-02-27", + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/global/gpt-4o-2024-11-20": { + "deprecation_date": "2026-03-01", + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/global/gpt-5.1": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/global/gpt-5.1-chat": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/global/gpt-5.1-codex": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/global/gpt-5.1-codex-mini": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 2e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-3.5-turbo": { + "input_cost_per_token": 5e-07, + "litellm_provider": "azure", + "max_input_tokens": 4097, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-3.5-turbo-0125": { + "deprecation_date": "2025-03-31", + "input_cost_per_token": 5e-07, + "litellm_provider": "azure", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-3.5-turbo-instruct-0914": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "azure_text", + "max_input_tokens": 4097, + "max_tokens": 4097, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "azure/gpt-35-turbo": { + "input_cost_per_token": 5e-07, + "litellm_provider": "azure", + "max_input_tokens": 4097, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-35-turbo-0125": { + "deprecation_date": "2025-05-31", + "input_cost_per_token": 5e-07, + "litellm_provider": "azure", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-35-turbo-1106": { + "deprecation_date": "2025-03-31", + "input_cost_per_token": 1e-06, + "litellm_provider": "azure", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-35-turbo-16k": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_tool_choice": true + }, + "azure/gpt-35-turbo-16k-0613": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-35-turbo-instruct": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "azure_text", + "max_input_tokens": 4097, + "max_tokens": 4097, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "azure/gpt-35-turbo-instruct-0914": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "azure_text", + "max_input_tokens": 4097, + "max_tokens": 4097, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "azure/gpt-4": { + "input_cost_per_token": 3e-05, + "litellm_provider": "azure", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-4-0125-preview": { + "input_cost_per_token": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-4-0613": { + "input_cost_per_token": 3e-05, + "litellm_provider": "azure", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-4-1106-preview": { + "input_cost_per_token": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-4-32k": { + "input_cost_per_token": 6e-05, + "litellm_provider": "azure", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.00012, + "supports_tool_choice": true + }, + "azure/gpt-4-32k-0613": { + "input_cost_per_token": 6e-05, + "litellm_provider": "azure", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.00012, + "supports_tool_choice": true + }, + "azure/gpt-4-turbo": { + "input_cost_per_token": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "azure/gpt-4-turbo-2024-04-09": { + "input_cost_per_token": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4-turbo-vision-preview": { + "input_cost_per_token": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4.1": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-06, + "output_cost_per_token_batches": 4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": false + }, + "azure/gpt-4.1-2025-04-14": { + "deprecation_date": "2026-11-04", + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-06, + "output_cost_per_token_batches": 4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": false + }, + "azure/gpt-4.1-mini": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 4e-07, + "input_cost_per_token_batches": 2e-07, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "output_cost_per_token_batches": 8e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": false + }, + "azure/gpt-4.1-mini-2025-04-14": { + "deprecation_date": "2026-11-04", + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 4e-07, + "input_cost_per_token_batches": 2e-07, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "output_cost_per_token_batches": 8e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": false + }, + "azure/gpt-4.1-nano": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 1e-07, + "input_cost_per_token_batches": 5e-08, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "output_cost_per_token_batches": 2e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4.1-nano-2025-04-14": { + "deprecation_date": "2026-11-04", + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 1e-07, + "input_cost_per_token_batches": 5e-08, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "output_cost_per_token_batches": 2e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4.5-preview": { + "cache_read_input_token_cost": 3.75e-05, + "input_cost_per_token": 7.5e-05, + "input_cost_per_token_batches": 3.75e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 0.00015, + "output_cost_per_token_batches": 7.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4o": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4o-2024-05-13": { + "input_cost_per_token": 5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4o-2024-08-06": { + "deprecation_date": "2026-02-27", + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4o-2024-11-20": { + "deprecation_date": "2026-03-01", + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-audio-2025-08-28": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/gpt-audio-1.5-2026-02-23": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/gpt-audio-mini-2025-10-06": { + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/gpt-4o-audio-preview-2024-12-17": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/gpt-4o-mini": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.65e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6.6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4o-mini-2024-07-18": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.65e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6.6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-4o-mini-audio-preview-2024-12-17": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/gpt-4o-mini-realtime-preview-2024-12-17": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/gpt-realtime-2025-08-28": { + "cache_creation_input_audio_token_cost": 4e-06, + "cache_read_input_token_cost": 4e-06, + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_image": 5e-06, + "input_cost_per_token": 4e-06, + "litellm_provider": "azure", + "max_input_tokens": 32000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1.6e-05, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/gpt-realtime-1.5-2026-02-23": { + "cache_creation_input_audio_token_cost": 4e-06, + "cache_read_input_token_cost": 4e-06, + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_image": 5e-06, + "input_cost_per_token": 4e-06, + "litellm_provider": "azure", + "max_input_tokens": 32000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1.6e-05, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/gpt-realtime-mini-2025-10-06": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_token_cost": 6e-08, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_image": 8e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "azure", + "max_input_tokens": 32000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/gpt-4o-mini-transcribe": { + "input_cost_per_audio_token": 3e-06, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 5e-06, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "azure/gpt-4o-mini-tts": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "mode": "audio_speech", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_second": 0.00025, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/speech" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "audio" + ] + }, + "azure/gpt-4o-realtime-preview-2024-10-01": { + "cache_creation_input_audio_token_cost": 2e-05, + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_audio_token": 0.0001, + "input_cost_per_token": 5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 0.0002, + "output_cost_per_token": 2e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/gpt-4o-realtime-preview-2024-12-17": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 2e-05, + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/gpt-4o-transcribe": { + "input_cost_per_audio_token": 6e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "azure/gpt-4o-transcribe-diarize": { + "input_cost_per_audio_token": 6e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "azure/gpt-5.1-2025-11-13": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_none_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "azure/gpt-5.1-chat-2025-11-13": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_native_streaming": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": false, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/gpt-5.1-codex-2025-11-13": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.1-codex-mini-2025-11-13": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_priority": 4.5e-08, + "input_cost_per_token": 2.5e-07, + "input_cost_per_token_priority": 4.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 2e-06, + "output_cost_per_token_priority": 3.6e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-2025-08-07": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-chat": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "source": "https://azure.microsoft.com/en-us/blog/gpt-5-in-azure-ai-foundry-the-future-of-ai-apps-and-agents-starts-here/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-chat-latest": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-codex": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-mini": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-mini-2025-08-07": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-nano": { + "cache_read_input_token_cost": 5e-09, + "input_cost_per_token": 5e-08, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-nano-2025-08-07": { + "cache_read_input_token_cost": 5e-09, + "input_cost_per_token": 5e-08, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5-pro": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00012, + "source": "https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-models/concepts/models-sold-directly-by-azure?pivots=azure-openai&tabs=global-standard-aoai%2Cstandard-chat-completions%2Cglobal-standard#gpt-5", + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.1": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/gpt-5.1-chat": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/gpt-5.1-codex": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.1-codex-max": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.1-codex-mini": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 2e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.2": { + "cache_read_input_token_cost": 1.75e-07, + "input_cost_per_token": 1.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.2-2025-12-11": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "azure/gpt-5.2-chat": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.2-chat-2025-12-11": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.2-codex": { + "cache_read_input_token_cost": 1.75e-07, + "input_cost_per_token": 1.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1.4e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.3-chat": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "azure/gpt-5.3-codex": { + "cache_read_input_token_cost": 1.75e-07, + "input_cost_per_token": 1.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1.4e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/gpt-5.2-pro": { + "input_cost_per_token": 2.1e-05, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.000168, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/gpt-5.2-pro-2025-12-11": { + "input_cost_per_token": 2.1e-05, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.000168, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/gpt-5.4": { + "cache_read_input_token_cost": 2.5e-07, + "cache_read_input_token_cost_above_272k_tokens": 5e-07, + "cache_read_input_token_cost_priority": 5e-07, + "cache_read_input_token_cost_above_272k_tokens_priority": 1e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_above_272k_tokens": 5e-06, + "input_cost_per_token_priority": 5e-06, + "input_cost_per_token_above_272k_tokens_priority": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_272k_tokens": 2.25e-05, + "output_cost_per_token_priority": 3e-05, + "output_cost_per_token_above_272k_tokens_priority": 4.5e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "azure/gpt-5.4-2026-03-05": { + "cache_read_input_token_cost": 2.5e-07, + "cache_read_input_token_cost_above_272k_tokens": 5e-07, + "cache_read_input_token_cost_priority": 5e-07, + "cache_read_input_token_cost_above_272k_tokens_priority": 1e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_above_272k_tokens": 5e-06, + "input_cost_per_token_priority": 5e-06, + "input_cost_per_token_above_272k_tokens_priority": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_272k_tokens": 2.25e-05, + "output_cost_per_token_priority": 3e-05, + "output_cost_per_token_above_272k_tokens_priority": 4.5e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "azure/gpt-5.4-pro": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/gpt-5.4-pro-2026-03-05": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/gpt-5.5": { + "cache_read_input_token_cost": 5e-07, + "cache_read_input_token_cost_above_272k_tokens": 1e-06, + "cache_read_input_token_cost_priority": 1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_token_above_272k_tokens": 1e-05, + "input_cost_per_token_priority": 1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-05, + "output_cost_per_token_above_272k_tokens": 4.5e-05, + "output_cost_per_token_priority": 6e-05, + "output_cost_per_token_above_272k_tokens_priority": 9e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": false + }, + "azure/gpt-5.5-2026-04-23": { + "cache_read_input_token_cost": 5e-07, + "cache_read_input_token_cost_above_272k_tokens": 1e-06, + "cache_read_input_token_cost_priority": 1e-06, + "cache_read_input_token_cost_above_272k_tokens_priority": 2e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_token_above_272k_tokens": 1e-05, + "input_cost_per_token_priority": 1e-05, + "input_cost_per_token_above_272k_tokens_priority": 2e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-05, + "output_cost_per_token_above_272k_tokens": 4.5e-05, + "output_cost_per_token_priority": 6e-05, + "output_cost_per_token_above_272k_tokens_priority": 9e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/gpt-5.5-pro": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": false, + "supports_low_reasoning_effort": false + }, + "azure/gpt-5.5-pro-2026-04-23": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/gpt-5.4-mini": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 7.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4.5e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false + }, + "azure/gpt-5.4-mini-2026-03-17": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 7.5e-07, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4.5e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false + }, + "azure/gpt-5.4-nano": { + "cache_read_input_token_cost": 2e-08, + "input_cost_per_token": 2e-07, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false + }, + "azure/gpt-5.4-nano-2026-03-17": { + "cache_read_input_token_cost": 2e-08, + "input_cost_per_token": 2e-07, + "litellm_provider": "azure", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false + }, + "azure/gpt-image-1": { + "cache_read_input_image_token_cost": 2.5e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_image_token": 1e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_image_token": 4e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "azure/hd/1024-x-1024/dall-e-3": { + "input_cost_per_pixel": 7.629e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/hd/1024-x-1792/dall-e-3": { + "input_cost_per_pixel": 6.539e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/hd/1792-x-1024/dall-e-3": { + "input_cost_per_pixel": 6.539e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/high/1024-x-1024/gpt-image-1": { + "input_cost_per_pixel": 1.59263611e-07, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/high/1024-x-1536/gpt-image-1": { + "input_cost_per_pixel": 1.58945719e-07, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/high/1536-x-1024/gpt-image-1": { + "input_cost_per_pixel": 1.58945719e-07, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/low/1024-x-1024/gpt-image-1": { + "input_cost_per_pixel": 1.0490417e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/low/1024-x-1536/gpt-image-1": { + "input_cost_per_pixel": 1.0172526e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/low/1536-x-1024/gpt-image-1": { + "input_cost_per_pixel": 1.0172526e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/medium/1024-x-1024/gpt-image-1": { + "input_cost_per_pixel": 4.0054321e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/medium/1024-x-1536/gpt-image-1": { + "input_cost_per_pixel": 4.0054321e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/medium/1536-x-1024/gpt-image-1": { + "input_cost_per_pixel": 4.0054321e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/gpt-image-1-mini": { + "cache_read_input_image_token_cost": 2.5e-07, + "cache_read_input_token_cost": 2e-07, + "input_cost_per_image_token": 2.5e-06, + "input_cost_per_token": 2e-06, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_image_token": 8e-06, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "azure/gpt-image-1.5": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_image_token": 8e-06, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_image_token": 3.2e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "azure/gpt-image-1.5-2025-12-16": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_image_token": 8e-06, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_image_token": 3.2e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "azure/gpt-image-2": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_image_token": 8e-06, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 1e-05, + "output_cost_per_image_token": 3e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "azure/gpt-image-2-2026-04-21": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_image_token": 8e-06, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 1e-05, + "output_cost_per_image_token": 3e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "azure/low/1024-x-1024/gpt-image-1-mini": { + "input_cost_per_pixel": 2.0751953125e-09, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/low/1024-x-1536/gpt-image-1-mini": { + "input_cost_per_pixel": 2.0751953125e-09, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/low/1536-x-1024/gpt-image-1-mini": { + "input_cost_per_pixel": 2.0345052083e-09, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/medium/1024-x-1024/gpt-image-1-mini": { + "input_cost_per_pixel": 8.056640625e-09, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/medium/1024-x-1536/gpt-image-1-mini": { + "input_cost_per_pixel": 8.056640625e-09, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/medium/1536-x-1024/gpt-image-1-mini": { + "input_cost_per_pixel": 7.9752604167e-09, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/high/1024-x-1024/gpt-image-1-mini": { + "input_cost_per_pixel": 3.173828125e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/high/1024-x-1536/gpt-image-1-mini": { + "input_cost_per_pixel": 3.173828125e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/high/1536-x-1024/gpt-image-1-mini": { + "input_cost_per_pixel": 3.1575520833e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure/mistral-large-2402": { + "input_cost_per_token": 8e-06, + "litellm_provider": "azure", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_function_calling": true + }, + "azure/mistral-large-latest": { + "input_cost_per_token": 8e-06, + "litellm_provider": "azure", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_function_calling": true + }, + "azure/o1": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o1-2024-12-17": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o1-mini": { + "cache_read_input_token_cost": 6.05e-07, + "input_cost_per_token": 1.21e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.84e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_vision": false + }, + "azure/o1-mini-2024-09-12": { + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_vision": false + }, + "azure/o1-preview": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_vision": false + }, + "azure/o1-preview-2024-09-12": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_vision": false + }, + "azure/o3": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o3-2025-04-16": { + "deprecation_date": "2026-04-16", + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o3-deep-research": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_token": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 4e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "azure/o3-mini": { + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/o3-mini-2025-01-31": { + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/o3-pro": { + "input_cost_per_token": 2e-05, + "input_cost_per_token_batches": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 8e-05, + "output_cost_per_token_batches": 4e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o3-pro-2025-06-10": { + "input_cost_per_token": 2e-05, + "input_cost_per_token_batches": 1e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 8e-05, + "output_cost_per_token_batches": 4e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o4-mini": { + "cache_read_input_token_cost": 2.75e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/o4-mini-2025-04-16": { + "cache_read_input_token_cost": 2.75e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/standard/1024-x-1024/dall-e-2": { + "input_cost_per_pixel": 0.0, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/standard/1024-x-1024/dall-e-3": { + "input_cost_per_pixel": 3.81469e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/standard/1024-x-1792/dall-e-3": { + "input_cost_per_pixel": 4.359e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/standard/1792-x-1024/dall-e-3": { + "input_cost_per_pixel": 4.359e-08, + "litellm_provider": "azure", + "mode": "image_generation", + "output_cost_per_token": 0.0 + }, + "azure/text-embedding-3-large": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "azure", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "azure/text-embedding-3-small": { + "deprecation_date": "2026-04-30", + "input_cost_per_token": 2e-08, + "litellm_provider": "azure", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "azure/text-embedding-ada-002": { + "input_cost_per_token": 1e-07, + "litellm_provider": "azure", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "azure/speech/azure-tts": { + "input_cost_per_character": 1.5e-05, + "litellm_provider": "azure", + "mode": "audio_speech", + "source": "https://azure.microsoft.com/en-us/pricing/calculator/" + }, + "azure/speech/azure-tts-hd": { + "input_cost_per_character": 3e-05, + "litellm_provider": "azure", + "mode": "audio_speech", + "source": "https://azure.microsoft.com/en-us/pricing/calculator/" + }, + "azure/tts-1": { + "input_cost_per_character": 1.5e-05, + "litellm_provider": "azure", + "mode": "audio_speech" + }, + "azure/tts-1-hd": { + "input_cost_per_character": 3e-05, + "litellm_provider": "azure", + "mode": "audio_speech" + }, + "azure/us/gpt-4.1-2025-04-14": { + "deprecation_date": "2026-11-04", + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 2.2e-06, + "input_cost_per_token_batches": 1.1e-06, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8.8e-06, + "output_cost_per_token_batches": 4.4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": false + }, + "azure/us/gpt-4.1-mini-2025-04-14": { + "deprecation_date": "2026-11-04", + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 4.4e-07, + "input_cost_per_token_batches": 2.2e-07, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.76e-06, + "output_cost_per_token_batches": 8.8e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": false + }, + "azure/us/gpt-4.1-nano-2025-04-14": { + "deprecation_date": "2026-11-04", + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 1.1e-07, + "input_cost_per_token_batches": 6e-08, + "litellm_provider": "azure", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4.4e-07, + "output_cost_per_token_batches": 2.2e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-4o-2024-08-06": { + "deprecation_date": "2026-02-27", + "cache_read_input_token_cost": 1.375e-06, + "input_cost_per_token": 2.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-4o-2024-11-20": { + "deprecation_date": "2026-03-01", + "cache_creation_input_token_cost": 1.38e-06, + "input_cost_per_token": 2.75e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-4o-mini-2024-07-18": { + "cache_read_input_token_cost": 8.3e-08, + "input_cost_per_token": 1.65e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6.6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-4o-mini-realtime-preview-2024-12-17": { + "cache_creation_input_audio_token_cost": 3.3e-07, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_audio_token": 1.1e-05, + "input_cost_per_token": 6.6e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2.2e-05, + "output_cost_per_token": 2.64e-06, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/us/gpt-4o-realtime-preview-2024-10-01": { + "cache_creation_input_audio_token_cost": 2.2e-05, + "cache_read_input_token_cost": 2.75e-06, + "input_cost_per_audio_token": 0.00011, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 0.00022, + "output_cost_per_token": 2.2e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/us/gpt-4o-realtime-preview-2024-12-17": { + "cache_read_input_audio_token_cost": 2.5e-06, + "cache_read_input_token_cost": 2.75e-06, + "input_cost_per_audio_token": 4.4e-05, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 2.2e-05, + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "azure/us/gpt-5-2025-08-07": { + "cache_read_input_token_cost": 1.375e-07, + "input_cost_per_token": 1.375e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-5-mini-2025-08-07": { + "cache_read_input_token_cost": 2.75e-08, + "input_cost_per_token": 2.75e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-5-nano-2025-08-07": { + "cache_read_input_token_cost": 5.5e-09, + "input_cost_per_token": 5.5e-08, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4.4e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-5.1": { + "cache_read_input_token_cost": 1.4e-07, + "input_cost_per_token": 1.38e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/us/gpt-5.1-chat": { + "cache_read_input_token_cost": 1.4e-07, + "input_cost_per_token": 1.38e-06, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_none_reasoning_effort": true + }, + "azure/us/gpt-5.1-codex": { + "cache_read_input_token_cost": 1.4e-07, + "input_cost_per_token": 1.38e-06, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1.1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/gpt-5.1-codex-mini": { + "cache_read_input_token_cost": 2.8e-08, + "input_cost_per_token": 2.75e-07, + "litellm_provider": "azure", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 2.2e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/o1-2024-12-17": { + "cache_read_input_token_cost": 8.25e-06, + "input_cost_per_token": 1.65e-05, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6.6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/o1-mini-2024-09-12": { + "cache_read_input_token_cost": 6.05e-07, + "input_cost_per_token": 1.21e-06, + "input_cost_per_token_batches": 6.05e-07, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.84e-06, + "output_cost_per_token_batches": 2.42e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_vision": false + }, + "azure/us/o1-preview-2024-09-12": { + "cache_read_input_token_cost": 8.25e-06, + "input_cost_per_token": 1.65e-05, + "litellm_provider": "azure", + "max_input_tokens": 128000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6.6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_vision": false + }, + "azure/us/o3-2025-04-16": { + "deprecation_date": "2026-04-16", + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 2.2e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8.8e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/us/o3-mini-2025-01-31": { + "cache_read_input_token_cost": 6.05e-07, + "input_cost_per_token": 1.21e-06, + "input_cost_per_token_batches": 6.05e-07, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.84e-06, + "output_cost_per_token_batches": 2.42e-06, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure/us/o4-mini-2025-04-16": { + "cache_read_input_token_cost": 3.1e-07, + "input_cost_per_token": 1.21e-06, + "litellm_provider": "azure", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.84e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure/whisper-1": { + "input_cost_per_second": 0.0001, + "litellm_provider": "azure", + "mode": "audio_transcription", + "output_cost_per_second": 0.0001 + }, + "azure_ai/Cohere-embed-v3-english": { + "input_cost_per_token": 1e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice", + "supports_embedding_image_input": true + }, + "azure_ai/Cohere-embed-v3-multilingual": { + "input_cost_per_token": 1e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/cohere.cohere-embed-v3-english-offer?tab=PlansAndPrice", + "supports_embedding_image_input": true + }, + "azure_ai/FLUX-1.1-pro": { + "litellm_provider": "azure_ai", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/black-forest-labs-flux-1-kontext-pro-and-flux1-1-pro-now-available-in-azure-ai-f/4434659", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure_ai/FLUX.1-Kontext-pro": { + "litellm_provider": "azure_ai", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure_ai/flux.2-pro": { + "litellm_provider": "azure_ai", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://ai.azure.com/explore/models/flux.2-pro/version/1/registry/azureml-blackforestlabs", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "azure_ai/Llama-3.2-11B-Vision-Instruct": { + "input_cost_per_token": 3.7e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 3.7e-07, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-11b-vision-instruct-offer?tab=Overview", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/Llama-3.2-90B-Vision-Instruct": { + "input_cost_per_token": 2.04e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 2.04e-06, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.meta-llama-3-2-90b-vision-instruct-offer?tab=Overview", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/Llama-3.3-70B-Instruct": { + "input_cost_per_token": 7.1e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 7.1e-07, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/metagenai.llama-3-3-70b-instruct-offer?tab=Overview", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/Llama-4-Maverick-17B-128E-Instruct-FP8": { + "input_cost_per_token": 1.41e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 1000000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 3.5e-07, + "source": "https://azure.microsoft.com/en-us/blog/introducing-the-llama-4-herd-in-azure-ai-foundry-and-azure-databricks/", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/Llama-4-Scout-17B-16E-Instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 10000000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 7.8e-07, + "source": "https://azure.microsoft.com/en-us/blog/introducing-the-llama-4-herd-in-azure-ai-foundry-and-azure-databricks/", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/Meta-Llama-3-70B-Instruct": { + "input_cost_per_token": 1.1e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 8192, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 3.7e-07, + "supports_tool_choice": true + }, + "azure_ai/Meta-Llama-3.1-405B-Instruct": { + "input_cost_per_token": 5.33e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 1.6e-05, + "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-405b-instruct-offer?tab=PlansAndPrice", + "supports_tool_choice": true + }, + "azure_ai/Meta-Llama-3.1-70B-Instruct": { + "input_cost_per_token": 2.68e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 3.54e-06, + "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-70b-instruct-offer?tab=PlansAndPrice", + "supports_tool_choice": true + }, + "azure_ai/Meta-Llama-3.1-8B-Instruct": { + "input_cost_per_token": 3e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 6.1e-07, + "source": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/metagenai.meta-llama-3-1-8b-instruct-offer?tab=PlansAndPrice", + "supports_tool_choice": true + }, + "azure_ai/Phi-3-medium-128k-instruct": { + "input_cost_per_token": 1.7e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6.8e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3-medium-4k-instruct": { + "input_cost_per_token": 1.7e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6.8e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3-mini-128k-instruct": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5.2e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3-mini-4k-instruct": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5.2e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3-small-128k-instruct": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3-small-8k-instruct": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3.5-MoE-instruct": { + "input_cost_per_token": 1.6e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6.4e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3.5-mini-instruct": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5.2e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-3.5-vision-instruct": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5.2e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/phi-3/", + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/Phi-4": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 5e-07, + "source": "https://techcommunity.microsoft.com/blog/machinelearningblog/affordable-innovation-unveiling-the-pricing-of-phi-3-slms-on-models-as-a-service/4156495", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "azure_ai/Phi-4-mini-instruct": { + "input_cost_per_token": 7.5e-08, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://techcommunity.microsoft.com/blog/Azure-AI-Services-blog/announcing-new-phi-pricing-empowering-your-business-with-small-language-models/4395112", + "supports_function_calling": true + }, + "azure_ai/Phi-4-multimodal-instruct": { + "input_cost_per_audio_token": 4e-06, + "input_cost_per_token": 8e-08, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3.2e-07, + "source": "https://techcommunity.microsoft.com/blog/Azure-AI-Services-blog/announcing-new-phi-pricing-empowering-your-business-with-small-language-models/4395112", + "supports_audio_input": true, + "supports_function_calling": true, + "supports_vision": true + }, + "azure_ai/Phi-4-mini-reasoning": { + "input_cost_per_token": 8e-08, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3.2e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/", + "supports_function_calling": true + }, + "azure_ai/Phi-4-reasoning": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5e-07, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true + }, + "azure_ai/mistral-document-ai-2505": { + "litellm_provider": "azure_ai", + "ocr_cost_per_page": 0.003, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://devblogs.microsoft.com/foundry/whats-new-in-azure-ai-foundry-august-2025/#mistral-document-ai-(ocr)-%E2%80%94-serverless-in-foundry" + }, + "azure_ai/mistral-document-ai-2512": { + "litellm_provider": "azure_ai", + "ocr_cost_per_page": 0.003, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/" + }, + "azure_ai/doc-intelligence/prebuilt-read": { + "litellm_provider": "azure_ai", + "ocr_cost_per_page": 0.0015, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-document-intelligence/" + }, + "azure_ai/doc-intelligence/prebuilt-layout": { + "litellm_provider": "azure_ai", + "ocr_cost_per_page": 0.01, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-document-intelligence/" + }, + "azure_ai/doc-intelligence/prebuilt-document": { + "litellm_provider": "azure_ai", + "ocr_cost_per_page": 0.01, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-document-intelligence/" + }, + "azure_ai/MAI-DS-R1": { + "input_cost_per_token": 1.35e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5.4e-06, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/microsoft/", + "supports_reasoning": true, + "supports_tool_choice": true + }, + "azure_ai/cohere-rerank-v3-english": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "azure_ai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "azure_ai/cohere-rerank-v3-multilingual": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "azure_ai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "azure_ai/cohere-rerank-v3.5": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "azure_ai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "azure_ai/cohere-rerank-v4.0-pro": { + "input_cost_per_query": 0.0025, + "input_cost_per_token": 0.0, + "litellm_provider": "azure_ai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_query_tokens": 4096, + "max_tokens": 32768, + "mode": "rerank", + "output_cost_per_token": 0.0, + "source": "https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-cohere-rerank-4-0-in-microsoft-foundry/4477076" + }, + "azure_ai/cohere-rerank-v4.0-fast": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "azure_ai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_query_tokens": 4096, + "max_tokens": 32768, + "mode": "rerank", + "output_cost_per_token": 0.0, + "source": "https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-cohere-rerank-4-0-in-microsoft-foundry/4477076" + }, + "azure_ai/deepseek-v3.2": { + "input_cost_per_token": 5.8e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "source": "https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-deepseek-v3-2-and-deepseek-v3-2-speciale-in-microsoft-foundry/4477549", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "azure_ai/deepseek-v3.2-speciale": { + "input_cost_per_token": 5.8e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "source": "https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/introducing-deepseek-v3-2-and-deepseek-v3-2-speciale-in-microsoft-foundry/4477549", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "azure_ai/deepseek-r1": { + "input_cost_per_token": 1.35e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5.4e-06, + "source": "https://techcommunity.microsoft.com/blog/machinelearningblog/deepseek-r1-improved-performance-higher-limits-and-transparent-pricing/4386367", + "supports_reasoning": true, + "supports_tool_choice": true + }, + "azure_ai/deepseek-v3": { + "input_cost_per_token": 1.14e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.56e-06, + "source": "https://techcommunity.microsoft.com/blog/machinelearningblog/announcing-deepseek-v3-on-azure-ai-foundry-and-github/4390438", + "supports_tool_choice": true + }, + "azure_ai/deepseek-v3-0324": { + "input_cost_per_token": 1.14e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.56e-06, + "source": "https://techcommunity.microsoft.com/blog/machinelearningblog/announcing-deepseek-v3-on-azure-ai-foundry-and-github/4390438", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/embed-v-4-0": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 3072, + "source": "https://azuremarketplace.microsoft.com/pt-br/marketplace/apps/cohere.cohere-embed-4-offer?tab=PlansAndPrice", + "supported_endpoints": [ + "/v1/embeddings" + ], + "supported_modalities": [ + "text", + "image" + ], + "supports_embedding_image_input": true + }, + "azure_ai/global/grok-3": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://devblogs.microsoft.com/foundry/announcing-grok-3-and-grok-3-mini-on-azure-ai-foundry/", + "supports_function_calling": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/global/grok-3-mini": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.27e-06, + "source": "https://devblogs.microsoft.com/foundry/announcing-grok-3-and-grok-3-mini-on-azure-ai-foundry/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-3": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/", + "supports_function_calling": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-3-mini": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.27e-06, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-4": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-4-fast-non-reasoning": { + "input_cost_per_token": 2e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-4-fast-reasoning": { + "input_cost_per_token": 2e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-4-1-fast-non-reasoning": { + "input_cost_per_token": 2e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "source": "https://techcommunity.microsoft.com/t5/Azure-AI-Foundry-Blog/Grok-4-0-Goes-GA-in-Microsoft-Foundry-and-Grok-4-1-Fast-Arrives/ba-p/4497964", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-4-1-fast-reasoning": { + "input_cost_per_token": 2e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "source": "https://techcommunity.microsoft.com/t5/Azure-AI-Foundry-Blog/Grok-4-0-Goes-GA-in-Microsoft-Foundry-and-Grok-4-1-Fast-Arrives/ba-p/4497964", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/grok-code-fast-1": { + "input_cost_per_token": 2e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/grok/", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "azure_ai/jais-30b-chat": { + "input_cost_per_token": 0.0032, + "litellm_provider": "azure_ai", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.00971, + "source": "https://azure.microsoft.com/en-us/products/ai-services/ai-foundry/models/jais-30b-chat" + }, + "azure_ai/jamba-instruct": { + "input_cost_per_token": 5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 70000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7e-07, + "supports_tool_choice": true + }, + "azure_ai/kimi-k2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://techcommunity.microsoft.com/blog/azure-ai-foundry-blog/kimi-k2-5-now-in-microsoft-foundry/4492321", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true + }, + "azure_ai/ministral-3b": { + "input_cost_per_token": 4e-08, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 4e-08, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.ministral-3b-2410-offer?tab=Overview", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/mistral-large": { + "input_cost_per_token": 4e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/mistral-large-2407": { + "input_cost_per_token": 2e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/mistral-large-latest": { + "input_cost_per_token": 2e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-ai-large-2407-offer?tab=Overview", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/mistral-large-3": { + "input_cost_per_token": 5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 256000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://azure.microsoft.com/en-us/blog/introducing-mistral-large-3-in-microsoft-foundry-open-capable-and-ready-for-production-workloads/", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "azure_ai/mistral-medium-2505": { + "input_cost_per_token": 4e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/mistral-nemo": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 131072, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://azuremarketplace.microsoft.com/en/marketplace/apps/000-000.mistral-nemo-12b-2407?tab=PlansAndPrice", + "supports_function_calling": true + }, + "azure_ai/mistral-small": { + "input_cost_per_token": 1e-06, + "litellm_provider": "azure_ai", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "azure_ai/mistral-small-2503": { + "input_cost_per_token": 1e-07, + "litellm_provider": "azure_ai", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "babbage-002": { + "input_cost_per_token": 4e-07, + "litellm_provider": "text-completion-openai", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 4e-07 + }, + "bedrock/*/1-month-commitment/cohere.command-light-text-v14": { + "input_cost_per_second": 0.001902, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_second": 0.001902, + "supports_tool_choice": true + }, + "bedrock/*/1-month-commitment/cohere.command-text-v14": { + "input_cost_per_second": 0.011, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_second": 0.011, + "supports_tool_choice": true + }, + "bedrock/*/6-month-commitment/cohere.command-light-text-v14": { + "input_cost_per_second": 0.0011416, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_second": 0.0011416, + "supports_tool_choice": true + }, + "bedrock/*/6-month-commitment/cohere.command-text-v14": { + "input_cost_per_second": 0.0066027, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_second": 0.0066027, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/1-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.01475, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.01475, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/1-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.0455, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0455 + }, + "bedrock/ap-northeast-1/1-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.0455, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0455, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/6-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.008194, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.008194, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/6-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.02527, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.02527 + }, + "bedrock/ap-northeast-1/6-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.02527, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.02527, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/anthropic.claude-instant-v1": { + "input_cost_per_token": 2.23e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 7.55e-06, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/anthropic.claude-v1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/anthropic.claude-v2:1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/ap-northeast-1/deepseek.v3.2": { + "input_cost_per_token": 7.4e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 2.22e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-northeast-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-northeast-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/ap-northeast-1/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 7.3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.03e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/ap-northeast-1/moonshotai.kimi-k2.5": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-northeast-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 7.3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.03e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/moonshotai.kimi-k2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.03e-06, + "source": "https://platform.moonshot.ai/docs/guide/kimi-k2-5-quickstart", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true + }, + "bedrock/ap-south-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 3.18e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.2e-06 + }, + "bedrock/ap-south-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 7.2e-07 + }, + "bedrock/ap-south-1/deepseek.v3.2": { + "input_cost_per_token": 7.4e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 2.22e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-south-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-south-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/ap-south-1/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 7.1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.94e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/ap-south-1/moonshotai.kimi-k2.5": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-south-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-southeast-2/minimax.minimax-m2.5": { + "input_cost_per_token": 3.09e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.236e-06 + }, + "bedrock/ap-southeast-3/deepseek.v3.2": { + "input_cost_per_token": 7.4e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 2.22e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-southeast-3/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-southeast-3/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/ap-southeast-3/moonshotai.kimi-k2.5": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ap-southeast-3/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/ca-central-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 3.05e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.03e-06 + }, + "bedrock/ca-central-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6.9e-07 + }, + "bedrock/eu-north-1/deepseek.v3.2": { + "input_cost_per_token": 7.4e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 2.22e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-north-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-north-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/eu-north-1/moonshotai.kimi-k2.5": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-central-1/1-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.01635, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.01635, + "supports_tool_choice": true + }, + "bedrock/eu-central-1/1-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.0415, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0415 + }, + "bedrock/eu-central-1/1-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.0415, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0415, + "supports_tool_choice": true + }, + "bedrock/eu-central-1/6-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.009083, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.009083, + "supports_tool_choice": true + }, + "bedrock/eu-central-1/6-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.02305, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.02305 + }, + "bedrock/eu-central-1/6-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.02305, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.02305, + "supports_tool_choice": true + }, + "bedrock/eu-central-1/anthropic.claude-instant-v1": { + "input_cost_per_token": 2.48e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 8.38e-06, + "supports_tool_choice": true + }, + "bedrock/eu-central-1/anthropic.claude-v1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05 + }, + "bedrock/eu-central-1/anthropic.claude-v2:1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/eu-central-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-central-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/eu-central-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-west-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 2.86e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.78e-06 + }, + "bedrock/eu-west-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6.5e-07 + }, + "bedrock/eu-west-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-west-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/eu-west-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-west-2/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 3.45e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.55e-06 + }, + "bedrock/eu-west-2/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3.9e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 7.8e-07 + }, + "bedrock/eu-west-2/minimax.minimax-m2.1": { + "input_cost_per_token": 4.7e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.86e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-west-2/minimax.minimax-m2.5": { + "input_cost_per_token": 4.7e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.86e-06 + }, + "bedrock/eu-west-2/qwen.qwen3-coder-next": { + "input_cost_per_token": 7.8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.86e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-west-3/mistral.mistral-7b-instruct-v0:2": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.6e-07, + "supports_tool_choice": true + }, + "bedrock/eu-west-3/mistral.mistral-large-2402-v1:0": { + "input_cost_per_token": 1.04e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3.12e-05, + "supports_function_calling": true + }, + "bedrock/eu-west-3/mistral.mixtral-8x7b-instruct-v0:1": { + "input_cost_per_token": 5.9e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 9.1e-07, + "supports_tool_choice": true + }, + "bedrock/eu-south-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/eu-south-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/eu-south-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/invoke/anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Anthropic via Invoke route does not currently support pdf input." + }, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "bedrock/sa-east-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 4.45e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5.88e-06 + }, + "bedrock/sa-east-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.01e-06 + }, + "bedrock/sa-east-1/deepseek.v3.2": { + "input_cost_per_token": 7.4e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 2.22e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/sa-east-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/sa-east-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.44e-06 + }, + "bedrock/sa-east-1/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 7.3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.03e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/sa-east-1/moonshotai.kimi-k2.5": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/sa-east-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.44e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-1/1-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.011, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.011, + "supports_tool_choice": true + }, + "bedrock/us-east-1/1-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.0175, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0175 + }, + "bedrock/us-east-1/1-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.0175, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0175, + "supports_tool_choice": true + }, + "bedrock/us-east-1/6-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.00611, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.00611, + "supports_tool_choice": true + }, + "bedrock/us-east-1/6-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.00972, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.00972 + }, + "bedrock/us-east-1/6-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.00972, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.00972, + "supports_tool_choice": true + }, + "bedrock/us-east-1/anthropic.claude-instant-v1": { + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-06, + "supports_tool_choice": true + }, + "bedrock/us-east-1/anthropic.claude-v1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/us-east-1/anthropic.claude-v2:1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/us-east-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 2.65e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.5e-06 + }, + "bedrock/us-east-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07 + }, + "bedrock/us-east-1/mistral.mistral-7b-instruct-v0:2": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_tool_choice": true + }, + "bedrock/us-east-1/mistral.mistral-large-2402-v1:0": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_function_calling": true + }, + "bedrock/us-east-1/mistral.mixtral-8x7b-instruct-v0:1": { + "input_cost_per_token": 4.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 7e-07, + "supports_tool_choice": true + }, + "bedrock/us-east-1/deepseek.v3.2": { + "input_cost_per_token": 6.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.85e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-1/minimax.minimax-m2.1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.2e-06 + }, + "bedrock/us-east-1/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/us-east-1/moonshotai.kimi-k2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-1/qwen.qwen3-coder-next": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-2/deepseek.v3.2": { + "input_cost_per_token": 6.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.85e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-2/minimax.minimax-m2.1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-2/minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.2e-06 + }, + "bedrock/us-east-2/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/us-east-2/moonshotai.kimi-k2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-2/qwen.qwen3-coder-next": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-gov-east-1/amazon.nova-pro-v1:0": { + "input_cost_per_token": 9.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.84e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "bedrock/us-gov-east-1/amazon.titan-embed-text-v1": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536 + }, + "bedrock/us-gov-east-1/amazon.titan-embed-text-v2:0": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024 + }, + "bedrock/us-gov-east-1/amazon.titan-text-express-v1": { + "input_cost_per_token": 1.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1.7e-06 + }, + "bedrock/us-gov-east-1/amazon.titan-text-lite-v1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 4e-07 + }, + "bedrock/us-gov-east-1/amazon.titan-text-premier-v1:0": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.5e-06 + }, + "bedrock/us-gov-east-1/anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3.6e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.8e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3.6e-07, + "cache_creation_input_token_cost": 4.5e-06 + }, + "bedrock/us-gov-east-1/anthropic.claude-3-haiku-20240307-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07 + }, + "bedrock/us-gov-east-1/anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_native_structured_output": true + }, + "bedrock/us-gov-east-1/claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_native_structured_output": true + }, + "bedrock/us-gov-east-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 2.65e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 3.5e-06, + "supports_pdf_input": true + }, + "bedrock/us-gov-east-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 2.65e-06, + "supports_pdf_input": true + }, + "bedrock/us-gov-west-1/amazon.nova-pro-v1:0": { + "input_cost_per_token": 9.6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.84e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "bedrock/us-gov-west-1/amazon.titan-embed-text-v1": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536 + }, + "bedrock/us-gov-west-1/amazon.titan-embed-text-v2:0": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024 + }, + "bedrock/us-gov-west-1/amazon.titan-text-express-v1": { + "input_cost_per_token": 1.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1.7e-06 + }, + "bedrock/us-gov-west-1/amazon.titan-text-lite-v1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 4e-07 + }, + "bedrock/us-gov-west-1/amazon.titan-text-premier-v1:0": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 42000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.5e-06 + }, + "bedrock/us-gov-west-1/anthropic.claude-3-7-sonnet-20250219-v1:0": { + "cache_creation_input_token_cost": 4.5e-06, + "cache_read_input_token_cost": 3.6e-07, + "input_cost_per_token": 3.6e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.8e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "bedrock/us-gov-west-1/anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3.6e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.8e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3.6e-07, + "cache_creation_input_token_cost": 4.5e-06 + }, + "bedrock/us-gov-west-1/anthropic.claude-3-haiku-20240307-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07 + }, + "bedrock/us-gov-west-1/anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_native_structured_output": true + }, + "bedrock/us-gov-west-1/claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_native_structured_output": true + }, + "bedrock/us-gov-west-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 2.65e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 3.5e-06, + "supports_pdf_input": true + }, + "bedrock/us-gov-west-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 2.65e-06, + "supports_pdf_input": true + }, + "bedrock/us-west-1/meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 2.65e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.5e-06 + }, + "bedrock/us-west-1/meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07 + }, + "bedrock/us-west-2/1-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.011, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.011, + "supports_tool_choice": true + }, + "bedrock/us-west-2/1-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.0175, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0175 + }, + "bedrock/us-west-2/1-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.0175, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.0175, + "supports_tool_choice": true + }, + "bedrock/us-west-2/6-month-commitment/anthropic.claude-instant-v1": { + "input_cost_per_second": 0.00611, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.00611, + "supports_tool_choice": true + }, + "bedrock/us-west-2/6-month-commitment/anthropic.claude-v1": { + "input_cost_per_second": 0.00972, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.00972 + }, + "bedrock/us-west-2/6-month-commitment/anthropic.claude-v2:1": { + "input_cost_per_second": 0.00972, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_second": 0.00972, + "supports_tool_choice": true + }, + "bedrock/us-west-2/anthropic.claude-instant-v1": { + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-06, + "supports_tool_choice": true + }, + "bedrock/us-west-2/anthropic.claude-v1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/us-west-2/anthropic.claude-v2:1": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 100000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true + }, + "bedrock/us-west-2/mistral.mistral-7b-instruct-v0:2": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_tool_choice": true + }, + "bedrock/us-west-2/mistral.mistral-large-2402-v1:0": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_function_calling": true + }, + "bedrock/us-west-2/mistral.mixtral-8x7b-instruct-v0:1": { + "input_cost_per_token": 4.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 7e-07, + "supports_tool_choice": true + }, + "bedrock/us-west-2/deepseek.v3.2": { + "input_cost_per_token": 6.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.85e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-west-2/minimax.minimax-m2.1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-west-2/minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "output_cost_per_token": 1.2e-06 + }, + "bedrock/us-west-2/moonshotai.kimi-k2-thinking": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_function_calling": true, + "supports_reasoning": true + }, + "bedrock/us-west-2/moonshotai.kimi-k2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-west-2/qwen.qwen3-coder-next": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us.anthropic.claude-3-5-haiku-20241022-v1:0": { + "cache_creation_input_token_cost": 1e-06, + "cache_read_input_token_cost": 8e-08, + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "black_forest_labs/flux-kontext-pro": { + "litellm_provider": "black_forest_labs", + "mode": "image_edit", + "output_cost_per_image": 0.04, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/edits", + "/v1/images/generations" + ] + }, + "black_forest_labs/flux-kontext-max": { + "litellm_provider": "black_forest_labs", + "mode": "image_edit", + "output_cost_per_image": 0.08, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/edits", + "/v1/images/generations" + ] + }, + "black_forest_labs/flux-pro-1.0-fill": { + "litellm_provider": "black_forest_labs", + "mode": "image_edit", + "output_cost_per_image": 0.05, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "black_forest_labs/flux-pro-1.0-expand": { + "litellm_provider": "black_forest_labs", + "mode": "image_edit", + "output_cost_per_image": 0.05, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "black_forest_labs/flux-pro-1.1": { + "litellm_provider": "black_forest_labs", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "black_forest_labs/flux-pro-1.1-ultra": { + "litellm_provider": "black_forest_labs", + "mode": "image_generation", + "output_cost_per_image": 0.06, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "black_forest_labs/flux-dev": { + "litellm_provider": "black_forest_labs", + "mode": "image_generation", + "output_cost_per_image": 0.025, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "black_forest_labs/flux-pro": { + "litellm_provider": "black_forest_labs", + "mode": "image_generation", + "output_cost_per_image": 0.05, + "source": "https://bfl.ai/pricing", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "cerebras/llama-3.3-70b": { + "input_cost_per_token": 8.5e-07, + "litellm_provider": "cerebras", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "cerebras/llama3.1-70b": { + "input_cost_per_token": 6e-07, + "litellm_provider": "cerebras", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "cerebras/llama3.1-8b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "cerebras", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "cerebras/gpt-oss-120b": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "cerebras", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 7.5e-07, + "source": "https://www.cerebras.ai/blog/openai-gpt-oss-120b-runs-fastest-on-cerebras", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "cerebras/qwen-3-32b": { + "input_cost_per_token": 4e-07, + "litellm_provider": "cerebras", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 8e-07, + "source": "https://inference-docs.cerebras.ai/support/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "cerebras/zai-glm-4.6": { + "deprecation_date": "2026-01-20", + "input_cost_per_token": 2.25e-06, + "litellm_provider": "cerebras", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "source": "https://www.cerebras.ai/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "cerebras/zai-glm-4.7": { + "input_cost_per_token": 2.25e-06, + "litellm_provider": "cerebras", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "source": "https://www.cerebras.ai/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "chatdolphin": { + "input_cost_per_token": 5e-07, + "litellm_provider": "nlp_cloud", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 5e-07 + }, + "chatgpt-4o-latest": { + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gpt-4o-transcribe-diarize": { + "input_cost_per_audio_token": 6e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "claude-haiku-4-5-20251001": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_creation_input_token_cost_above_1hr": 2e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_computer_use": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "claude-haiku-4-5": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_creation_input_token_cost_above_1hr": 2e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_computer_use": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "claude-3-7-sonnet-20250219": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "deprecation_date": "2026-02-19", + "input_cost_per_token": 3e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-3-haiku-20240307": { + "cache_creation_input_token_cost": 3e-07, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 264 + }, + "claude-3-opus-20240229": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 1.5e-06, + "deprecation_date": "2026-05-01", + "input_cost_per_token": 1.5e-05, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 395 + }, + "claude-4-opus-20250514": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-4-sonnet-20250514": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost": 3e-07, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-sonnet-4-5": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346 + }, + "claude-sonnet-4-5-20250929": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tool_use_system_prompt_tokens": 346 + }, + "claude-sonnet-4-6": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_minimal_reasoning_effort": true + }, + "claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-opus-4-1": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_creation_input_token_cost_above_1hr": 3e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-opus-4-1-20250805": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_creation_input_token_cost_above_1hr": 3e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "deprecation_date": "2026-08-05", + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-opus-4-20250514": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_creation_input_token_cost_above_1hr": 3e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "deprecation_date": "2026-05-14", + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-opus-4-5-20251101": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-opus-4-5": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "claude-opus-4-6": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "provider_specific_entry": { + "us": 1.1, + "fast": 6.0 + }, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "claude-opus-4-6-20260205": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "provider_specific_entry": { + "us": 1.1, + "fast": 6.0 + }, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "claude-opus-4-7": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "supports_max_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "provider_specific_entry": { + "us": 1.1, + "fast": 6.0 + }, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "claude-opus-4-7-20260416": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "supports_max_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "provider_specific_entry": { + "us": 1.1, + "fast": 6.0 + }, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "claude-sonnet-4-20250514": { + "deprecation_date": "2026-05-14", + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "anthropic", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "cloudflare/@cf/meta/llama-2-7b-chat-fp16": { + "input_cost_per_token": 1.923e-06, + "litellm_provider": "cloudflare", + "max_input_tokens": 3072, + "max_output_tokens": 3072, + "max_tokens": 3072, + "mode": "chat", + "output_cost_per_token": 1.923e-06 + }, + "cloudflare/@cf/meta/llama-2-7b-chat-int8": { + "input_cost_per_token": 1.923e-06, + "litellm_provider": "cloudflare", + "max_input_tokens": 2048, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 1.923e-06 + }, + "cloudflare/@cf/mistral/mistral-7b-instruct-v0.1": { + "input_cost_per_token": 1.923e-06, + "litellm_provider": "cloudflare", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.923e-06 + }, + "cloudflare/@hf/thebloke/codellama-7b-instruct-awq": { + "input_cost_per_token": 1.923e-06, + "litellm_provider": "cloudflare", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.923e-06 + }, + "codestral/codestral-2405": { + "input_cost_per_token": 0.0, + "litellm_provider": "codestral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://docs.mistral.ai/capabilities/code_generation/", + "supports_assistant_prefill": true, + "supports_tool_choice": true + }, + "codestral/codestral-latest": { + "input_cost_per_token": 0.0, + "litellm_provider": "codestral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://docs.mistral.ai/capabilities/code_generation/", + "supports_assistant_prefill": true, + "supports_tool_choice": true + }, + "codex-mini-latest": { + "cache_read_input_token_cost": 3.75e-07, + "input_cost_per_token": 1.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 6e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "cohere.command-light-text-v14": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_tool_choice": true + }, + "cohere.command-r-plus-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_tool_choice": true + }, + "cohere.command-r-v1:0": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_tool_choice": true + }, + "cohere.command-text-v14": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_tool_choice": true + }, + "cohere.embed-english-v3": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "supports_embedding_image_input": true + }, + "cohere.embed-multilingual-v3": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "supports_embedding_image_input": true + }, + "cohere.embed-v4:0": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536, + "supports_embedding_image_input": true + }, + "cohere/embed-v4.0": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "cohere", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536, + "supports_embedding_image_input": true + }, + "cohere.rerank-v3-5:0": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "bedrock", + "max_document_chunks_per_query": 100, + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_query_tokens": 32000, + "max_tokens": 32000, + "max_tokens_per_document_chunk": 512, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "command": { + "input_cost_per_token": 1e-06, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "command-a-03-2025": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "cohere_chat", + "max_input_tokens": 256000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "command-light": { + "input_cost_per_token": 3e-07, + "litellm_provider": "cohere_chat", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_tool_choice": true + }, + "command-nightly": { + "input_cost_per_token": 1e-06, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "command-r": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "cohere_chat", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "command-r-08-2024": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "cohere_chat", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "command-r-plus": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "cohere_chat", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "command-r-plus-08-2024": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "cohere_chat", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "command-r7b-12-2024": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "cohere_chat", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3.75e-08, + "source": "https://docs.cohere.com/v2/docs/command-r7b", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "computer-use-preview": { + "input_cost_per_token": 3e-06, + "litellm_provider": "azure", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "dall-e-2": { + "input_cost_per_image": 0.02, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits", + "/v1/images/variations" + ] + }, + "dall-e-3": { + "input_cost_per_image": 0.04, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "deepseek-chat": { + "cache_read_input_token_cost": 2.8e-08, + "input_cost_per_token": 2.8e-07, + "litellm_provider": "deepseek", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.2e-07, + "source": "https://api-docs.deepseek.com/quick_start/pricing", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "deepseek-reasoner": { + "cache_read_input_token_cost": 2.8e-08, + "input_cost_per_token": 2.8e-07, + "litellm_provider": "deepseek", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.2e-07, + "source": "https://api-docs.deepseek.com/quick_start/pricing", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": false, + "supports_native_streaming": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": false + }, + "dashscope/qwen-coder": { + "input_cost_per_token": 3e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 1000000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-flash": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 5e-08, + "output_cost_per_token": 4e-07, + "range": [ + 0, + 256000.0 + ] + }, + { + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 2e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen-flash-2025-07-28": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 5e-08, + "output_cost_per_token": 4e-07, + "range": [ + 0, + 256000.0 + ] + }, + { + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 2e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen-max": { + "input_cost_per_token": 1.6e-06, + "litellm_provider": "dashscope", + "max_input_tokens": 30720, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6.4e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-plus": { + "input_cost_per_token": 4e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 129024, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-plus-2025-01-25": { + "input_cost_per_token": 4e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 129024, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-plus-2025-04-28": { + "input_cost_per_token": 4e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 129024, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-06, + "output_cost_per_token": 1.2e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-plus-2025-07-14": { + "input_cost_per_token": 4e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 129024, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-06, + "output_cost_per_token": 1.2e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-plus-2025-07-28": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 4e-07, + "output_cost_per_reasoning_token": 4e-06, + "output_cost_per_token": 1.2e-06, + "range": [ + 0, + 256000.0 + ] + }, + { + "input_cost_per_token": 1.2e-06, + "output_cost_per_reasoning_token": 1.2e-05, + "output_cost_per_token": 3.6e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen-plus-2025-09-11": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 4e-07, + "output_cost_per_reasoning_token": 4e-06, + "output_cost_per_token": 1.2e-06, + "range": [ + 0, + 256000.0 + ] + }, + { + "input_cost_per_token": 1.2e-06, + "output_cost_per_reasoning_token": 1.2e-05, + "output_cost_per_token": 3.6e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen-plus-latest": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 4e-07, + "output_cost_per_reasoning_token": 4e-06, + "output_cost_per_token": 1.2e-06, + "range": [ + 0, + 256000.0 + ] + }, + { + "input_cost_per_token": 1.2e-06, + "output_cost_per_reasoning_token": 1.2e-05, + "output_cost_per_token": 3.6e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen-turbo": { + "input_cost_per_token": 5e-08, + "litellm_provider": "dashscope", + "max_input_tokens": 129024, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_reasoning_token": 5e-07, + "output_cost_per_token": 2e-07, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-turbo-2024-11-01": { + "input_cost_per_token": 5e-08, + "litellm_provider": "dashscope", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-turbo-2025-04-28": { + "input_cost_per_token": 5e-08, + "litellm_provider": "dashscope", + "max_input_tokens": 1000000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_reasoning_token": 5e-07, + "output_cost_per_token": 2e-07, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-turbo-latest": { + "input_cost_per_token": 5e-08, + "litellm_provider": "dashscope", + "max_input_tokens": 1000000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_reasoning_token": 5e-07, + "output_cost_per_token": 2e-07, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen3-30b-a3b": { + "litellm_provider": "dashscope", + "max_input_tokens": 129024, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen3-coder-flash": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "cache_read_input_token_cost": 8e-08, + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.5e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "cache_read_input_token_cost": 1.2e-07, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 2.5e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 8e-07, + "output_cost_per_token": 4e-06, + "range": [ + 128000.0, + 256000.0 + ] + }, + { + "cache_read_input_token_cost": 4e-07, + "input_cost_per_token": 1.6e-06, + "output_cost_per_token": 9.6e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen3-coder-flash-2025-07-28": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.5e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 5e-07, + "output_cost_per_token": 2.5e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 8e-07, + "output_cost_per_token": 4e-06, + "range": [ + 128000.0, + 256000.0 + ] + }, + { + "input_cost_per_token": 1.6e-06, + "output_cost_per_token": 9.6e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen3-coder-plus": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 5e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "cache_read_input_token_cost": 1.8e-07, + "input_cost_per_token": 1.8e-06, + "output_cost_per_token": 9e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "range": [ + 128000.0, + 256000.0 + ] + }, + { + "cache_read_input_token_cost": 6e-07, + "input_cost_per_token": 6e-06, + "output_cost_per_token": 6e-05, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen3-coder-plus-2025-07-22": { + "litellm_provider": "dashscope", + "max_input_tokens": 997952, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 5e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 1.8e-06, + "output_cost_per_token": 9e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "range": [ + 128000.0, + 256000.0 + ] + }, + { + "input_cost_per_token": 6e-06, + "output_cost_per_token": 6e-05, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwen3-max-preview": { + "litellm_provider": "dashscope", + "max_input_tokens": 258048, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 6e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 2.4e-06, + "output_cost_per_token": 1.2e-05, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "range": [ + 128000.0, + 252000.0 + ] + } + ] + }, + "dashscope/qwen3-max": { + "litellm_provider": "dashscope", + "max_input_tokens": 258048, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 6e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 2.4e-06, + "output_cost_per_token": 1.2e-05, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "range": [ + 128000.0, + 252000.0 + ] + } + ] + }, + "dashscope/qwen3-max-2026-01-23": { + "litellm_provider": "dashscope", + "max_input_tokens": 258048, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "tiered_pricing": [ + { + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 6e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 2.4e-06, + "output_cost_per_token": 1.2e-05, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "range": [ + 128000.0, + 252000.0 + ] + } + ] + }, + "dashscope/qwen3-next-80b-a3b-instruct": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "dashscope/qwen3-next-80b-a3b-thinking": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen3-vl-235b-a22b-instruct": { + "input_cost_per_token": 4e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "dashscope/qwen3-vl-235b-a22b-thinking": { + "input_cost_per_token": 4e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "dashscope/qwen3-vl-32b-instruct": { + "input_cost_per_token": 1.6e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6.4e-07, + "source": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "dashscope/qwen3-vl-32b-thinking": { + "input_cost_per_token": 1.6e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 2.87e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/model-pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "dashscope/qwen3-vl-plus": { + "litellm_provider": "dashscope", + "max_input_tokens": 260096, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tiered_pricing": [ + { + "input_cost_per_token": 2e-07, + "output_cost_per_token": 1.6e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 2.4e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 6e-07, + "output_cost_per_token": 4.8e-06, + "range": [ + 128000.0, + 256000.0 + ] + } + ] + }, + "dashscope/qwen3.5-plus": { + "litellm_provider": "dashscope", + "max_input_tokens": 991808, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tiered_pricing": [ + { + "input_cost_per_token": 4e-07, + "output_cost_per_token": 2.4e-06, + "range": [ + 0, + 256000.0 + ] + }, + { + "input_cost_per_token": 5e-07, + "output_cost_per_token": 3e-06, + "range": [ + 256000.0, + 1000000.0 + ] + } + ] + }, + "dashscope/qwq-plus": { + "input_cost_per_token": 8e-07, + "litellm_provider": "dashscope", + "max_input_tokens": 98304, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.4e-06, + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "dashscope/qwen-image-2.0": { + "litellm_provider": "dashscope", + "mode": "image_generation", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "dashscope/qwen-image-2.0-pro": { + "litellm_provider": "dashscope", + "mode": "image_generation", + "source": "https://www.alibabacloud.com/help/en/model-studio/models", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "databricks/databricks-bge-large-en": { + "input_cost_per_token": 1.0003e-07, + "input_dbu_cost_per_token": 1.429e-06, + "litellm_provider": "databricks", + "max_input_tokens": 512, + "max_tokens": 512, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_dbu_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving" + }, + "databricks/databricks-claude-3-7-sonnet": { + "input_cost_per_token": 2.9999900000000002e-06, + "input_dbu_cost_per_token": 4.2857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000020000000002e-05, + "output_dbu_cost_per_token": 0.000214286, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-haiku-4-5": { + "input_cost_per_token": 1.00002e-06, + "input_dbu_cost_per_token": 1.4286e-05, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 5.00003e-06, + "output_dbu_cost_per_token": 7.1429e-05, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-opus-4": { + "input_cost_per_token": 1.5000020000000002e-05, + "input_dbu_cost_per_token": 0.000214286, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 7.500003000000001e-05, + "output_dbu_cost_per_token": 0.001071429, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-opus-4-1": { + "input_cost_per_token": 1.5000020000000002e-05, + "input_dbu_cost_per_token": 0.000214286, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 7.500003000000001e-05, + "output_dbu_cost_per_token": 0.001071429, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-opus-4-5": { + "input_cost_per_token": 5.00003e-06, + "input_dbu_cost_per_token": 7.1429e-05, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 2.5000010000000002e-05, + "output_dbu_cost_per_token": 0.000357143, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-sonnet-4": { + "input_cost_per_token": 2.9999900000000002e-06, + "input_dbu_cost_per_token": 4.2857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000020000000002e-05, + "output_dbu_cost_per_token": 0.000214286, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-sonnet-4-1": { + "input_cost_per_token": 2.9999900000000002e-06, + "input_dbu_cost_per_token": 4.2857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000020000000002e-05, + "output_dbu_cost_per_token": 0.000214286, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-claude-sonnet-4-5": { + "input_cost_per_token": 2.9999900000000002e-06, + "input_dbu_cost_per_token": 4.2857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000020000000002e-05, + "output_dbu_cost_per_token": 0.000214286, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "databricks/databricks-gemini-2-5-flash": { + "input_cost_per_token": 3.0001999999999996e-07, + "input_dbu_cost_per_token": 4.285999999999999e-06, + "litellm_provider": "databricks", + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_tokens": 65535, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 2.49998e-06, + "output_dbu_cost_per_token": 3.5714e-05, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "databricks/databricks-gemini-2-5-pro": { + "input_cost_per_token": 1.24999e-06, + "input_dbu_cost_per_token": 1.7857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 9.999990000000002e-06, + "output_dbu_cost_per_token": 0.000142857, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "databricks/databricks-gemma-3-12b": { + "input_cost_per_token": 1.5000999999999998e-07, + "input_dbu_cost_per_token": 2.1429999999999996e-06, + "litellm_provider": "databricks", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 5.0001e-07, + "output_dbu_cost_per_token": 7.143e-06, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving" + }, + "databricks/databricks-gpt-5": { + "input_cost_per_token": 1.24999e-06, + "input_dbu_cost_per_token": 1.7857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 9.999990000000002e-06, + "output_dbu_cost_per_token": 0.000142857, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving" + }, + "databricks/databricks-gpt-5-1": { + "input_cost_per_token": 1.24999e-06, + "input_dbu_cost_per_token": 1.7857e-05, + "litellm_provider": "databricks", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 9.999990000000002e-06, + "output_dbu_cost_per_token": 0.000142857, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving" + }, + "databricks/databricks-gpt-5-mini": { + "input_cost_per_token": 2.4997000000000006e-07, + "input_dbu_cost_per_token": 3.571e-06, + "litellm_provider": "databricks", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.9999700000000004e-06, + "output_dbu_cost_per_token": 2.8571e-05, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving" + }, + "databricks/databricks-gpt-5-nano": { + "input_cost_per_token": 4.998e-08, + "input_dbu_cost_per_token": 7.14e-07, + "litellm_provider": "databricks", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 3.9998000000000007e-07, + "output_dbu_cost_per_token": 5.714000000000001e-06, + "source": "https://www.databricks.com/product/pricing/proprietary-foundation-model-serving" + }, + "databricks/databricks-gpt-oss-120b": { + "input_cost_per_token": 1.5000999999999998e-07, + "input_dbu_cost_per_token": 2.1429999999999996e-06, + "litellm_provider": "databricks", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 5.9997e-07, + "output_dbu_cost_per_token": 8.571e-06, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving" + }, + "databricks/databricks-gpt-oss-20b": { + "input_cost_per_token": 7e-08, + "input_dbu_cost_per_token": 1e-06, + "litellm_provider": "databricks", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 3.0001999999999996e-07, + "output_dbu_cost_per_token": 4.285999999999999e-06, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving" + }, + "databricks/databricks-gte-large-en": { + "input_cost_per_token": 1.2999000000000001e-07, + "input_dbu_cost_per_token": 1.857e-06, + "litellm_provider": "databricks", + "max_input_tokens": 8192, + "max_tokens": 8192, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_dbu_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving" + }, + "databricks/databricks-llama-2-70b-chat": { + "input_cost_per_token": 5.0001e-07, + "input_dbu_cost_per_token": 7.143e-06, + "litellm_provider": "databricks", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000300000000002e-06, + "output_dbu_cost_per_token": 2.1429e-05, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-llama-4-maverick": { + "input_cost_per_token": 5.0001e-07, + "input_dbu_cost_per_token": 7.143e-06, + "litellm_provider": "databricks", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Databricks documentation now provides both DBU costs (_dbu_cost_per_token) and dollar costs(_cost_per_token)." + }, + "mode": "chat", + "output_cost_per_token": 1.5000300000000002e-06, + "output_dbu_cost_per_token": 2.1429e-05, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-meta-llama-3-1-405b-instruct": { + "input_cost_per_token": 5.00003e-06, + "input_dbu_cost_per_token": 7.1429e-05, + "litellm_provider": "databricks", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000020000000002e-05, + "output_dbu_cost_per_token": 0.000214286, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-meta-llama-3-1-8b-instruct": { + "input_cost_per_token": 1.5000999999999998e-07, + "input_dbu_cost_per_token": 2.1429999999999996e-06, + "litellm_provider": "databricks", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 4.5003000000000007e-07, + "output_dbu_cost_per_token": 6.429000000000001e-06, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving" + }, + "databricks/databricks-meta-llama-3-3-70b-instruct": { + "input_cost_per_token": 5.0001e-07, + "input_dbu_cost_per_token": 7.143e-06, + "litellm_provider": "databricks", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.5000300000000002e-06, + "output_dbu_cost_per_token": 2.1429e-05, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-meta-llama-3-70b-instruct": { + "input_cost_per_token": 1.00002e-06, + "input_dbu_cost_per_token": 1.4286e-05, + "litellm_provider": "databricks", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 2.9999900000000002e-06, + "output_dbu_cost_per_token": 4.2857e-05, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-mixtral-8x7b-instruct": { + "input_cost_per_token": 5.0001e-07, + "input_dbu_cost_per_token": 7.143e-06, + "litellm_provider": "databricks", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.00002e-06, + "output_dbu_cost_per_token": 1.4286e-05, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-mpt-30b-instruct": { + "input_cost_per_token": 1.00002e-06, + "input_dbu_cost_per_token": 1.4286e-05, + "litellm_provider": "databricks", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 1.00002e-06, + "output_dbu_cost_per_token": 1.4286e-05, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "databricks/databricks-mpt-7b-instruct": { + "input_cost_per_token": 5.0001e-07, + "input_dbu_cost_per_token": 7.143e-06, + "litellm_provider": "databricks", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "metadata": { + "notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation." + }, + "mode": "chat", + "output_cost_per_token": 0.0, + "output_dbu_cost_per_token": 0.0, + "source": "https://www.databricks.com/product/pricing/foundation-model-serving", + "supports_tool_choice": true + }, + "dataforseo/search": { + "input_cost_per_query": 0.003, + "litellm_provider": "dataforseo", + "mode": "search" + }, + "davinci-002": { + "input_cost_per_token": 2e-06, + "litellm_provider": "text-completion-openai", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "deepgram/base": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-conversationalai": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-finance": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-general": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-meeting": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-phonecall": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-video": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/base-voicemail": { + "input_cost_per_second": 0.00020833, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0125/60 seconds = $0.00020833 per second", + "original_pricing_per_minute": 0.0125 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/enhanced": { + "input_cost_per_second": 0.00024167, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0145/60 seconds = $0.00024167 per second", + "original_pricing_per_minute": 0.0145 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/enhanced-finance": { + "input_cost_per_second": 0.00024167, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0145/60 seconds = $0.00024167 per second", + "original_pricing_per_minute": 0.0145 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/enhanced-general": { + "input_cost_per_second": 0.00024167, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0145/60 seconds = $0.00024167 per second", + "original_pricing_per_minute": 0.0145 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/enhanced-meeting": { + "input_cost_per_second": 0.00024167, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0145/60 seconds = $0.00024167 per second", + "original_pricing_per_minute": 0.0145 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/enhanced-phonecall": { + "input_cost_per_second": 0.00024167, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0145/60 seconds = $0.00024167 per second", + "original_pricing_per_minute": 0.0145 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-atc": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-automotive": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-conversationalai": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-drivethru": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-finance": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-general": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-meeting": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-phonecall": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-video": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-2-voicemail": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-3": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-3-general": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-3-medical": { + "input_cost_per_second": 8.667e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0052/60 seconds = $0.00008667 per second (multilingual)", + "original_pricing_per_minute": 0.0052 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-general": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/nova-phonecall": { + "input_cost_per_second": 7.167e-05, + "litellm_provider": "deepgram", + "metadata": { + "calculation": "$0.0043/60 seconds = $0.00007167 per second", + "original_pricing_per_minute": 0.0043 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/whisper": { + "input_cost_per_second": 0.0001, + "litellm_provider": "deepgram", + "metadata": { + "notes": "Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models" + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/whisper-base": { + "input_cost_per_second": 0.0001, + "litellm_provider": "deepgram", + "metadata": { + "notes": "Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models" + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/whisper-large": { + "input_cost_per_second": 0.0001, + "litellm_provider": "deepgram", + "metadata": { + "notes": "Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models" + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/whisper-medium": { + "input_cost_per_second": 0.0001, + "litellm_provider": "deepgram", + "metadata": { + "notes": "Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models" + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/whisper-small": { + "input_cost_per_second": 0.0001, + "litellm_provider": "deepgram", + "metadata": { + "notes": "Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models" + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepgram/whisper-tiny": { + "input_cost_per_second": 0.0001, + "litellm_provider": "deepgram", + "metadata": { + "notes": "Deepgram's hosted OpenAI Whisper models - pricing may differ from native Deepgram models" + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://deepgram.com/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "deepinfra/Gryphe/MythoMax-L2-13b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 9e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/NousResearch/Hermes-3-Llama-3.1-405B": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 1e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/NousResearch/Hermes-3-Llama-3.1-70B": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 3e-07, + "output_cost_per_token": 3e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/Qwen/QwQ-32B": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen2.5-72B-Instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1.2e-07, + "output_cost_per_token": 3.9e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen2.5-7B-Instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/Qwen/Qwen2.5-VL-32B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_vision": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-14B": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 2.4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-235B-A22B": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 1.8e-07, + "output_cost_per_token": 5.4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-235B-A22B-Instruct-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 9e-08, + "output_cost_per_token": 6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-235B-A22B-Thinking-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 3e-07, + "output_cost_per_token": 2.9e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-30B-A3B": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 2.9e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-32B": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 2.8e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 4e-07, + "output_cost_per_token": 1.6e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2.9e-07, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-Next-80B-A3B-Instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 1.4e-07, + "output_cost_per_token": 1.4e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Qwen/Qwen3-Next-80B-A3B-Thinking": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 1.4e-07, + "output_cost_per_token": 1.4e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/Sao10K/L3-8B-Lunaris-v1-Turbo": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 5e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/Sao10K/L3.1-70B-Euryale-v2.2": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 6.5e-07, + "output_cost_per_token": 7.5e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/Sao10K/L3.3-70B-Euryale-v2.3": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 6.5e-07, + "output_cost_per_token": 7.5e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/allenai/olmOCR-7B-0725-FP8": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1.5e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/anthropic/claude-3-7-sonnet-latest": { + "max_tokens": 200000, + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "input_cost_per_token": 3.3e-06, + "output_cost_per_token": 1.65e-05, + "cache_read_input_token_cost": 3.3e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/anthropic/claude-4-opus": { + "max_tokens": 200000, + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "input_cost_per_token": 1.65e-05, + "output_cost_per_token": 8.25e-05, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/anthropic/claude-4-sonnet": { + "max_tokens": 200000, + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "input_cost_per_token": 3.3e-06, + "output_cost_per_token": 1.65e-05, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-R1": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 7e-07, + "output_cost_per_token": 2.4e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-R1-0528": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 2.15e-06, + "cache_read_input_token_cost": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-R1-0528-Turbo": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-R1-Distill-Llama-70B": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 2.7e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-R1-Turbo": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-V3": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 3.8e-07, + "output_cost_per_token": 8.9e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-V3-0324": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 8.8e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-V3.1": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1e-06, + "cache_read_input_token_cost": 2.16e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_function_calling": true + }, + "deepinfra/deepseek-ai/DeepSeek-V3.1-Terminus": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1e-06, + "cache_read_input_token_cost": 2.16e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/google/gemini-2.0-flash-001": { + "deprecation_date": "2026-06-01", + "max_tokens": 1000000, + "max_input_tokens": 1000000, + "max_output_tokens": 1000000, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/google/gemini-2.5-flash": { + "max_tokens": 1000000, + "max_input_tokens": 1000000, + "max_output_tokens": 1000000, + "input_cost_per_token": 3e-07, + "output_cost_per_token": 2.5e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/google/gemini-2.5-pro": { + "max_tokens": 1000000, + "max_input_tokens": 1000000, + "max_output_tokens": 1000000, + "input_cost_per_token": 1.25e-06, + "output_cost_per_token": 1e-05, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/google/gemma-3-12b-it": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 5e-08, + "output_cost_per_token": 1e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/google/gemma-3-27b-it": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-08, + "output_cost_per_token": 1.6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/google/gemma-3-4b-it": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 8e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Llama-3.2-11B-Vision-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 4.9e-08, + "output_cost_per_token": 4.9e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/meta-llama/Llama-3.2-3B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 2e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Llama-3.3-70B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2.3e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Llama-3.3-70B-Instruct-Turbo": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 3.9e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "deepinfra/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8": { + "max_tokens": 1048576, + "max_input_tokens": 1048576, + "max_output_tokens": 1048576, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Llama-4-Scout-17B-16E-Instruct": { + "max_tokens": 327680, + "max_input_tokens": 327680, + "max_output_tokens": 327680, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 3e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Llama-Guard-3-8B": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 5.5e-08, + "output_cost_per_token": 5.5e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/meta-llama/Llama-Guard-4-12B": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 1.8e-07, + "output_cost_per_token": 1.8e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/meta-llama/Meta-Llama-3-8B-Instruct": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 3e-08, + "output_cost_per_token": 6e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Meta-Llama-3.1-70B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 4e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 2.8e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Meta-Llama-3.1-8B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 3e-08, + "output_cost_per_token": 5e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 3e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/microsoft/WizardLM-2-8x22B": { + "max_tokens": 65536, + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "input_cost_per_token": 4.8e-07, + "output_cost_per_token": 4.8e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": false + }, + "deepinfra/microsoft/phi-4": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 7e-08, + "output_cost_per_token": 1.4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/mistralai/Mistral-Nemo-Instruct-2407": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 4e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/mistralai/Mistral-Small-24B-Instruct-2501": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-08, + "output_cost_per_token": 8e-08, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/mistralai/Mistral-Small-3.2-24B-Instruct-2506": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 7.5e-08, + "output_cost_per_token": 2e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/mistralai/Mixtral-8x7B-Instruct-v0.1": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 4e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/moonshotai/Kimi-K2-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 2e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/moonshotai/Kimi-K2-Instruct-0905": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 2e-06, + "cache_read_input_token_cost": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/nvidia/Llama-3.1-Nemotron-70B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/nvidia/Llama-3.3-Nemotron-Super-49B-v1.5": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/nvidia/NVIDIA-Nemotron-Nano-9B-v2": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1.6e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/openai/gpt-oss-120b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 5e-08, + "output_cost_per_token": 4.5e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/openai/gpt-oss-20b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1.5e-07, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepinfra/zai-org/GLM-4.5": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 4e-07, + "output_cost_per_token": 1.6e-06, + "litellm_provider": "deepinfra", + "mode": "chat", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "deepseek/deepseek-chat": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 2.8e-08, + "input_cost_per_token": 2.8e-07, + "input_cost_per_token_cache_hit": 2.8e-08, + "litellm_provider": "deepseek", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4.2e-07, + "source": "https://api-docs.deepseek.com/quick_start/pricing", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "deepseek/deepseek-coder": { + "input_cost_per_token": 1.4e-07, + "input_cost_per_token_cache_hit": 1.4e-08, + "litellm_provider": "deepseek", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true + }, + "deepseek/deepseek-r1": { + "input_cost_per_token": 5.5e-07, + "input_cost_per_token_cache_hit": 1.4e-07, + "litellm_provider": "deepseek", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.19e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "deepseek/deepseek-reasoner": { + "cache_read_input_token_cost": 2.8e-08, + "input_cost_per_token": 2.8e-07, + "input_cost_per_token_cache_hit": 2.8e-08, + "litellm_provider": "deepseek", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.2e-07, + "source": "https://api-docs.deepseek.com/quick_start/pricing", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_assistant_prefill": true, + "supports_function_calling": false, + "supports_native_streaming": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": false + }, + "deepseek/deepseek-v3": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 7e-08, + "input_cost_per_token": 2.7e-07, + "input_cost_per_token_cache_hit": 7e-08, + "litellm_provider": "deepseek", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.1e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true + }, + "deepseek/deepseek-v3.2": { + "input_cost_per_token": 2.8e-07, + "input_cost_per_token_cache_hit": 2.8e-08, + "litellm_provider": "deepseek", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "deepseek.v3-v1:0": { + "input_cost_per_token": 5.8e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 163840, + "max_output_tokens": 81920, + "max_tokens": 81920, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_native_structured_output": true + }, + "deepseek.v3.2": { + "input_cost_per_token": 6.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.85e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "dolphin": { + "input_cost_per_token": 5e-07, + "litellm_provider": "nlp_cloud", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "completion", + "output_cost_per_token": 5e-07 + }, + "deepseek-v3-2-251201": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 98304, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "glm-4-7-251222": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 204800, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "kimi-k2-thinking-251104": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 229376, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "doubao-embedding": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Volcengine Doubao embedding model - standard version with 2560 dimensions" + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 2560 + }, + "doubao-embedding-large": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Volcengine Doubao embedding model - large version with 2048 dimensions" + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 2048 + }, + "doubao-embedding-large-text-240915": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Volcengine Doubao embedding model - text-240915 version with 4096 dimensions" + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 4096 + }, + "doubao-embedding-large-text-250515": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Volcengine Doubao embedding model - text-250515 version with 2048 dimensions" + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 2048 + }, + "doubao-embedding-text-240715": { + "input_cost_per_token": 0.0, + "litellm_provider": "volcengine", + "max_input_tokens": 4096, + "max_tokens": 4096, + "metadata": { + "notes": "Volcengine Doubao embedding model - text-240715 version with 2560 dimensions" + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 2560 + }, + "exa_ai/search": { + "litellm_provider": "exa_ai", + "mode": "search", + "tiered_pricing": [ + { + "input_cost_per_query": 0.005, + "max_results_range": [ + 0, + 25 + ] + }, + { + "input_cost_per_query": 0.025, + "max_results_range": [ + 26, + 100 + ] + } + ] + }, + "firecrawl/search": { + "litellm_provider": "firecrawl", + "mode": "search", + "tiered_pricing": [ + { + "input_cost_per_query": 0.00166, + "max_results_range": [ + 1, + 10 + ] + }, + { + "input_cost_per_query": 0.00332, + "max_results_range": [ + 11, + 20 + ] + }, + { + "input_cost_per_query": 0.00498, + "max_results_range": [ + 21, + 30 + ] + }, + { + "input_cost_per_query": 0.00664, + "max_results_range": [ + 31, + 40 + ] + }, + { + "input_cost_per_query": 0.0083, + "max_results_range": [ + 41, + 50 + ] + }, + { + "input_cost_per_query": 0.00996, + "max_results_range": [ + 51, + 60 + ] + }, + { + "input_cost_per_query": 0.01162, + "max_results_range": [ + 61, + 70 + ] + }, + { + "input_cost_per_query": 0.01328, + "max_results_range": [ + 71, + 80 + ] + }, + { + "input_cost_per_query": 0.01494, + "max_results_range": [ + 81, + 90 + ] + }, + { + "input_cost_per_query": 0.0166, + "max_results_range": [ + 91, + 100 + ] + } + ], + "metadata": { + "notes": "Firecrawl search pricing: $83 for 100,000 credits, 2 credits per 10 results. Cost = ceiling(limit/10) * 2 * $0.00083" + } + }, + "perplexity/search": { + "input_cost_per_query": 0.005, + "litellm_provider": "perplexity", + "mode": "search" + }, + "searxng/search": { + "litellm_provider": "searxng", + "mode": "search", + "input_cost_per_query": 0.0, + "metadata": { + "notes": "SearXNG is an open-source metasearch engine. Free to use when self-hosted or using public instances." + } + }, + "serper/search": { + "input_cost_per_query": 0.001, + "litellm_provider": "serper", + "mode": "search", + "metadata": { + "notes": "Serper Google Search API. Pricing: $1.00/1k queries (Starter), $0.75/1k (Standard), $0.50/1k (Scale), $0.30/1k (Ultimate)." + } + }, + "elevenlabs/scribe_v1": { + "input_cost_per_second": 6.11e-05, + "litellm_provider": "elevenlabs", + "metadata": { + "calculation": "$0.22/hour = $0.00366/minute = $0.0000611 per second (enterprise pricing)", + "notes": "ElevenLabs Scribe v1 - state-of-the-art speech recognition model with 99 language support", + "original_pricing_per_hour": 0.22 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://elevenlabs.io/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "elevenlabs/scribe_v1_experimental": { + "input_cost_per_second": 6.11e-05, + "litellm_provider": "elevenlabs", + "metadata": { + "calculation": "$0.22/hour = $0.00366/minute = $0.0000611 per second (enterprise pricing)", + "notes": "ElevenLabs Scribe v1 experimental - enhanced version of the main Scribe model", + "original_pricing_per_hour": 0.22 + }, + "mode": "audio_transcription", + "output_cost_per_second": 0.0, + "source": "https://elevenlabs.io/pricing", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "elevenlabs/eleven_v3": { + "input_cost_per_character": 0.00018, + "litellm_provider": "elevenlabs", + "metadata": { + "calculation": "$0.18/1000 characters (Scale plan pricing, 1 credit per character)", + "notes": "ElevenLabs Eleven v3 - most expressive TTS model with 70+ languages and audio tags support" + }, + "mode": "audio_speech", + "source": "https://elevenlabs.io/pricing", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "elevenlabs/eleven_multilingual_v2": { + "input_cost_per_character": 0.00018, + "litellm_provider": "elevenlabs", + "metadata": { + "calculation": "$0.18/1000 characters (Scale plan pricing, 1 credit per character)", + "notes": "ElevenLabs Eleven Multilingual v2 - default TTS model with 29 languages support" + }, + "mode": "audio_speech", + "source": "https://elevenlabs.io/pricing", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "embed-english-light-v2.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "cohere", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "embed-english-light-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "cohere", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "embed-english-v2.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_tokens": 4096, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "embed-english-v3.0": { + "input_cost_per_image": 0.0001, + "input_cost_per_token": 1e-07, + "litellm_provider": "cohere", + "max_input_tokens": 1024, + "max_tokens": 1024, + "metadata": { + "notes": "'supports_image_input' is a deprecated field. Use 'supports_embedding_image_input' instead." + }, + "mode": "embedding", + "output_cost_per_token": 0.0, + "supports_embedding_image_input": true, + "supports_image_input": true + }, + "embed-multilingual-v2.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "cohere", + "max_input_tokens": 768, + "max_tokens": 768, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "embed-multilingual-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "cohere", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "embedding", + "output_cost_per_token": 0.0, + "supports_embedding_image_input": true + }, + "embed-multilingual-light-v3.0": { + "input_cost_per_token": 0.0001, + "litellm_provider": "cohere", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "embedding", + "output_cost_per_token": 0.0, + "supports_embedding_image_input": true + }, + "eu.amazon.nova-lite-v1:0": { + "input_cost_per_token": 7.8e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.12e-07, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "eu.amazon.nova-micro-v1:0": { + "input_cost_per_token": 4.6e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.84e-07, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true + }, + "eu.amazon.nova-pro-v1:0": { + "input_cost_per_token": 1.05e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 4.2e-06, + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "eu.anthropic.claude-3-5-haiku-20241022-v1:0": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "cache_read_input_token_cost": 2.5e-08, + "cache_creation_input_token_cost": 3.125e-07 + }, + "eu.anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.375e-06, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 1.1e-06, + "deprecation_date": "2026-10-15", + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5.5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "eu.anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "eu.anthropic.claude-3-5-sonnet-20241022-v2:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "eu.anthropic.claude-3-7-sonnet-20250219-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "eu.anthropic.claude-3-haiku-20240307-v1:0": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 2.5e-08, + "cache_creation_input_token_cost": 3.125e-07 + }, + "eu.anthropic.claude-3-opus-20240229-v1:0": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 1.5e-06, + "cache_creation_input_token_cost": 1.875e-05 + }, + "eu.anthropic.claude-3-sonnet-20240229-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "eu.anthropic.claude-opus-4-1-20250805-v1:0": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "eu.anthropic.claude-opus-4-20250514-v1:0": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "eu.anthropic.claude-sonnet-4-20250514-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "eu.anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "input_cost_per_token_above_200k_tokens": 6.6e-06, + "output_cost_per_token_above_200k_tokens": 2.475e-05, + "cache_creation_input_token_cost_above_200k_tokens": 8.25e-06, + "cache_read_input_token_cost_above_200k_tokens": 6.6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "eu.meta.llama3-2-1b-instruct-v1:0": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.3e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "eu.meta.llama3-2-3b-instruct-v1:0": { + "input_cost_per_token": 1.9e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.9e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "eu.mistral.pixtral-large-2502-v1:0": { + "input_cost_per_token": 2e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "fal_ai/bria/text-to-image/3.2": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.0398, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/flux-pro/v1.1": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/flux-pro/v1.1-ultra": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.06, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/flux/schnell": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.003, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/bytedance/seedream/v3/text-to-image": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.03, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/bytedance/dreamina/v3.1/text-to-image": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.03, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/ideogram/v3": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.06, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/imagen4/preview": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.0398, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/imagen4/preview/fast": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.02, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/imagen4/preview/ultra": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.06, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/recraft/v3/text-to-image": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.0398, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "fal_ai/fal-ai/stable-diffusion-v35-medium": { + "litellm_provider": "fal_ai", + "mode": "image_generation", + "output_cost_per_image": 0.0398, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "featherless_ai/featherless-ai/Qwerky-72B": { + "litellm_provider": "featherless_ai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat" + }, + "featherless_ai/featherless-ai/Qwerky-QwQ-32B": { + "litellm_provider": "featherless_ai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat" + }, + "fireworks-ai-4.1b-to-16b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "output_cost_per_token": 2e-07 + }, + "fireworks-ai-56b-to-176b": { + "input_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "output_cost_per_token": 1.2e-06 + }, + "fireworks-ai-above-16b": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "output_cost_per_token": 9e-07 + }, + "fireworks-ai-default": { + "input_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "output_cost_per_token": 0.0 + }, + "fireworks-ai-embedding-150m-to-350m": { + "input_cost_per_token": 1.6e-08, + "litellm_provider": "fireworks_ai-embedding-models", + "output_cost_per_token": 0.0 + }, + "fireworks-ai-embedding-up-to-150m": { + "input_cost_per_token": 8e-09, + "litellm_provider": "fireworks_ai-embedding-models", + "output_cost_per_token": 0.0 + }, + "fireworks-ai-moe-up-to-56b": { + "input_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "output_cost_per_token": 5e-07 + }, + "fireworks-ai-up-to-4b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "output_cost_per_token": 2e-07 + }, + "fireworks_ai/WhereIsAI/UAE-Large-V1": { + "input_cost_per_token": 1.6e-08, + "litellm_provider": "fireworks_ai-embedding-models", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "source": "https://fireworks.ai/pricing" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-v2-instruct": { + "input_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1": { + "input_cost_per_token": 3e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 20480, + "max_tokens": 20480, + "mode": "chat", + "output_cost_per_token": 8e-06, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-0528": { + "input_cost_per_token": 3e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 160000, + "max_output_tokens": 160000, + "max_tokens": 160000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-basic": { + "input_cost_per_token": 5.5e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 20480, + "max_tokens": 20480, + "mode": "chat", + "output_cost_per_token": 2.19e-06, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v3": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v3-0324": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://fireworks.ai/models/fireworks/deepseek-v3-0324", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v3p1": { + "input_cost_per_token": 5.6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "source": "https://fireworks.ai/pricing", + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v3p1-terminus": { + "input_cost_per_token": 5.6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "source": "https://fireworks.ai/pricing", + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v3p2": { + "input_cost_per_token": 5.6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "source": "https://fireworks.ai/models/fireworks/deepseek-v3p2", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/firefunction-v2": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/glm-4p5": { + "input_cost_per_token": 5.5e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 96000, + "max_tokens": 96000, + "mode": "chat", + "output_cost_per_token": 2.19e-06, + "source": "https://fireworks.ai/models/fireworks/glm-4p5", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/glm-4p5-air": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 96000, + "max_tokens": 96000, + "mode": "chat", + "output_cost_per_token": 8.8e-07, + "source": "https://artificialanalysis.ai/models/glm-4-5-air", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/glm-4p6": { + "input_cost_per_token": 5.5e-07, + "output_cost_per_token": 2.19e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 202800, + "max_output_tokens": 202800, + "max_tokens": 202800, + "mode": "chat", + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/glm-4p7": { + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 202800, + "max_output_tokens": 202800, + "max_tokens": 202800, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "source": "https://fireworks.ai/models/fireworks/glm-4p7", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/gpt-oss-120b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/gpt-oss-20b": { + "input_cost_per_token": 5e-08, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/kimi-k2-instruct": { + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://fireworks.ai/models/fireworks/kimi-k2-instruct", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/kimi-k2-instruct-0905": { + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 262144, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://app.fireworks.ai/models/fireworks/kimi-k2-instruct-0905", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/kimi-k2-thinking": { + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "fireworks_ai/accounts/fireworks/models/kimi-k2p5": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p1-405b-instruct": { + "input_cost_per_token": 3e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p1-8b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p2-11b-vision-instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p2-1b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p2-3b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p2-90b-vision-instruct": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "fireworks_ai/accounts/fireworks/models/llama4-maverick-instruct-basic": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 8.8e-07, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/llama4-scout-instruct-basic": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://fireworks.ai/pricing", + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/minimax-m2p1": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_token": 3e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 204800, + "max_output_tokens": 204800, + "max_tokens": 204800, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://fireworks.ai/models/fireworks/minimax-m2p1", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/mixtral-8x22b-instruct-hf": { + "input_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/accounts/fireworks/models/qwen2-72b-instruct": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct": { + "input_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/accounts/fireworks/models/yi-large": { + "input_cost_per_token": 3e-06, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "fireworks_ai/glm-4p7": { + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 202800, + "max_output_tokens": 202800, + "max_tokens": 202800, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "source": "https://fireworks.ai/models/fireworks/glm-4p7", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/kimi-k2p5": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://fireworks.ai/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/minimax-m2p1": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_token": 3e-07, + "litellm_provider": "fireworks_ai", + "max_input_tokens": 204800, + "max_output_tokens": 204800, + "max_tokens": 204800, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://fireworks.ai/models/fireworks/minimax-m2p1", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "fireworks_ai/nomic-ai/nomic-embed-text-v1": { + "input_cost_per_token": 8e-09, + "litellm_provider": "fireworks_ai-embedding-models", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "source": "https://fireworks.ai/pricing" + }, + "fireworks_ai/nomic-ai/nomic-embed-text-v1.5": { + "input_cost_per_token": 8e-09, + "litellm_provider": "fireworks_ai-embedding-models", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0.0, + "source": "https://fireworks.ai/pricing" + }, + "fireworks_ai/thenlper/gte-base": { + "input_cost_per_token": 8e-09, + "litellm_provider": "fireworks_ai-embedding-models", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "source": "https://fireworks.ai/pricing" + }, + "fireworks_ai/thenlper/gte-large": { + "input_cost_per_token": 1.6e-08, + "litellm_provider": "fireworks_ai-embedding-models", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "source": "https://fireworks.ai/pricing" + }, + "friendliai/meta-llama-3.1-70b-instruct": { + "input_cost_per_token": 6e-07, + "litellm_provider": "friendliai", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "friendliai/meta-llama-3.1-8b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "friendliai", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:babbage-002": { + "input_cost_per_token": 1.6e-06, + "input_cost_per_token_batches": 2e-07, + "litellm_provider": "text-completion-openai", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 1.6e-06, + "output_cost_per_token_batches": 2e-07 + }, + "ft:davinci-002": { + "input_cost_per_token": 1.2e-05, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "text-completion-openai", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_batches": 1e-06 + }, + "ft:gpt-3.5-turbo": { + "input_cost_per_token": 3e-06, + "input_cost_per_token_batches": 1.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "output_cost_per_token_batches": 3e-06, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-3.5-turbo-0125": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-3.5-turbo-0613": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openai", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-3.5-turbo-1106": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-4-0613": { + "input_cost_per_token": 3e-05, + "litellm_provider": "openai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "source": "OpenAI needs to add pricing for this ft model, will be updated when added by OpenAI. Defaulting to base model pricing", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-4o-2024-08-06": { + "cache_read_input_token_cost": 1.875e-06, + "input_cost_per_token": 3.75e-06, + "input_cost_per_token_batches": 1.875e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_batches": 7.5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "ft:gpt-4o-2024-11-20": { + "cache_creation_input_token_cost": 1.875e-06, + "input_cost_per_token": 3.75e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-4o-mini-2024-07-18": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 3e-07, + "input_cost_per_token_batches": 1.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "output_cost_per_token_batches": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-4.1-2025-04-14": { + "cache_read_input_token_cost": 7.5e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_batches": 1.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_batches": 6e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-4.1-mini-2025-04-14": { + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 8e-07, + "input_cost_per_token_batches": 4e-07, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "output_cost_per_token_batches": 1.6e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:gpt-4.1-nano-2025-04-14": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_batches": 1e-07, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-07, + "output_cost_per_token_batches": 4e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "ft:o4-mini-2025-04-16": { + "cache_read_input_token_cost": 1e-06, + "input_cost_per_token": 4e-06, + "input_cost_per_token_batches": 2e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 1.6e-05, + "output_cost_per_token_batches": 8e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "gemini-2.0-flash": { + "cache_read_input_token_cost": 2.5e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 4e-07, + "source": "https://ai.google.dev/pricing#2_0flash", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.0-flash-001": { + "cache_read_input_token_cost": 3.75e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.0-flash-lite": { + "cache_read_input_token_cost": 1.875e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7.5e-08, + "input_cost_per_token": 7.5e-08, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 50, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.0-flash-lite-001": { + "cache_read_input_token_cost": 1.875e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7.5e-08, + "input_cost_per_token": 7.5e-08, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 50, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#gemini-2.0-flash", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.5-flash": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + }, + "supports_service_tier": true + }, + "gemini-2.5-flash-image": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "max_pdf_size_mb": 30, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "image_generation", + "output_cost_per_image": 0.039, + "output_cost_per_image_token": 3e-05, + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 100000, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-flash-image", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": false, + "tpm": 8000000, + "supports_service_tier": true + }, + "gemini-3-pro-image-preview": { + "input_cost_per_image": 0.0011, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.134, + "output_cost_per_image_token": 0.00012, + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_batches": 6e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query", + "supports_service_tier": true + }, + "gemini-3.1-flash-image-preview": { + "input_cost_per_image": 0.00056, + "input_cost_per_token": 5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.0672, + "output_cost_per_image_token": 6e-05, + "output_cost_per_token": 3e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini-3.1-flash-lite-preview": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_per_audio_token": 5e-08, + "input_cost_per_audio_token": 5e-07, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 1.5e-06, + "output_cost_per_token": 1.5e-06, + "source": "https://ai.google.dev/gemini-api/docs/models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_audio_output": false, + "supports_code_execution": true, + "supports_file_search": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query", + "supports_service_tier": true + }, + "deep-research-pro-preview-12-2025": { + "input_cost_per_image": 0.0011, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.134, + "output_cost_per_image_token": 0.00012, + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_batches": 6e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true + }, + "gemini-2.5-flash-lite": { + "cache_read_input_token_cost": 1e-08, + "input_cost_per_audio_token": 3e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + }, + "supports_service_tier": true + }, + "gemini-2.5-flash-lite-preview-09-2025": { + "cache_read_input_token_cost": 1e-08, + "input_cost_per_audio_token": 3e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "source": "https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.5-flash-preview-09-2025": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "source": "https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-live-2.5-flash-preview-native-audio-09-2025": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_audio_token": 3e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "realtime", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_token": 2e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/vertex_ai/live" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-live-2.5-flash-preview-native-audio-09-2025": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_audio_token": 3e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "realtime", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_token": 2e-06, + "rpm": 100000, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 8000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.5-flash-lite-preview-06-17": { + "deprecation_date": "2025-11-18", + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_audio_token": 5e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.5-pro": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_above_200k_tokens": 2.5e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + }, + "supports_service_tier": true + }, + "gemini-3-pro-preview": { + "deprecation_date": "2026-03-26", + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini-3.1-pro-preview": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "output_cost_per_image": 0.00012, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_url_context": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini-3.1-pro-preview-customtools": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "output_cost_per_image": 0.00012, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_url_context": true, + "supports_native_streaming": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "vertex_ai/gemini-3-pro-preview": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "vertex_ai/gemini-3-flash-preview": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 5e-07, + "input_cost_per_audio_token": 1e-06, + "litellm_provider": "vertex_ai", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 9e-07, + "input_cost_per_audio_token_priority": 1.8e-06, + "output_cost_per_token_priority": 5.4e-06, + "cache_read_input_token_cost_priority": 9e-08, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "vertex_ai/gemini-3.1-pro-preview": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "output_cost_per_image": 0.00012, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_url_context": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "vertex_ai/gemini-3.1-pro-preview-customtools": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "output_cost_per_image": 0.00012, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_url_context": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini-2.5-pro-preview-tts": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "audio" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-robotics-er-1.5-preview": { + "cache_read_input_token_cost": 0, + "input_cost_per_token": 3e-07, + "input_cost_per_audio_token": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_tokens": 65535, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "output_cost_per_reasoning_token": 2.5e-06, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-robotics-er-1-5-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image", + "video", + "audio" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true + }, + "gemini/gemini-robotics-er-1.5-preview": { + "cache_read_input_token_cost": 0, + "input_cost_per_token": 3e-07, + "input_cost_per_audio_token": 1e-06, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_tokens": 65535, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "output_cost_per_reasoning_token": 2.5e-06, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-robotics-er-1-5-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image", + "video", + "audio" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "rpm": 10, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.5-computer-use-preview-10-2025": { + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "vertex_ai-language-models", + "max_images_per_prompt": 3000, + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/computer-use", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_computer_use": true, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gemini-embedding-001": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models" + }, + "gemini-embedding-2-preview": { + "input_cost_per_audio_per_second": 0.00016, + "input_cost_per_image": 0.00012, + "input_cost_per_token": 2e-07, + "input_cost_per_video_per_second": 0.00079, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "uses_embed_content": true + }, + "gemini-embedding-2": { + "input_cost_per_audio_per_second": 0.00016, + "input_cost_per_image": 0.00012, + "input_cost_per_token": 2e-07, + "input_cost_per_video_per_second": 0.00079, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supports_multimodal": true, + "uses_embed_content": true + }, + "vertex_ai/gemini-embedding-2-preview": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://ai.google.dev/gemini-api/docs/embeddings#multimodal", + "supports_multimodal": true, + "uses_embed_content": true + }, + "vertex_ai/gemini-embedding-2": { + "input_cost_per_audio_per_second": 0.00016, + "input_cost_per_image": 0.00012, + "input_cost_per_token": 2e-07, + "input_cost_per_video_per_second": 0.00079, + "litellm_provider": "vertex_ai", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://ai.google.dev/gemini-api/docs/embeddings#multimodal", + "supports_multimodal": true, + "uses_embed_content": true + }, + "gemini-flash-experimental": { + "input_cost_per_character": 0, + "input_cost_per_token": 0, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "uses_embed_content": true + }, + "gemini/gemini-embedding-001": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "gemini", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "rpm": 10000, + "source": "https://ai.google.dev/gemini-api/docs/embeddings#model-versions", + "tpm": 10000000 + }, + "gemini/gemini-embedding-2-preview": { + "input_cost_per_audio_per_second": 0.00016, + "input_cost_per_image": 0.00012, + "input_cost_per_token": 2e-07, + "input_cost_per_video_per_second": 0.00079, + "litellm_provider": "gemini", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "rpm": 10000, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supports_multimodal": true, + "tpm": 10000000 + }, + "gemini/gemini-embedding-2": { + "input_cost_per_audio_per_second": 0.00016, + "input_cost_per_image": 0.00012, + "input_cost_per_token": 2e-07, + "input_cost_per_video_per_second": 0.00079, + "litellm_provider": "gemini", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "rpm": 10000, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supports_multimodal": true, + "tpm": 10000000 + }, + "gemini/gemini-1.5-flash": { + "deprecation_date": "2025-09-29", + "input_cost_per_token": 7.5e-08, + "input_cost_per_token_above_128k_tokens": 1.5e-07, + "litellm_provider": "gemini", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "rpm": 10000, + "source": "https://ai.google.dev/gemini-api/docs/embeddings#multimodal", + "supports_multimodal": true, + "tpm": 10000000 + }, + "gemini/gemini-2.0-flash": { + "cache_read_input_token_cost": 2.5e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 4e-07, + "rpm": 10000, + "source": "https://ai.google.dev/pricing#2_0flash", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 10000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.0-flash-001": { + "cache_read_input_token_cost": 2.5e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 4e-07, + "rpm": 10000, + "source": "https://ai.google.dev/pricing#2_0flash", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 10000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.0-flash-lite": { + "cache_read_input_token_cost": 1.875e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7.5e-08, + "input_cost_per_token": 7.5e-08, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 50, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 3e-07, + "rpm": 4000, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-2.0-flash-lite", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 4000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.5-flash": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 100000, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 8000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + }, + "supports_service_tier": true + }, + "gemini/gemini-2.5-flash-image": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "supports_reasoning": false, + "max_images_per_prompt": 3000, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "max_pdf_size_mb": 30, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "image_generation", + "output_cost_per_image": 0.039, + "output_cost_per_image_token": 3e-05, + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 100000, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-flash-image", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 8000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + }, + "supports_service_tier": true + }, + "gemini/gemini-3-pro-image-preview": { + "input_cost_per_image": 0.0011, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "gemini", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.134, + "output_cost_per_image_token": 0.00012, + "output_cost_per_token": 1.2e-05, + "rpm": 1000, + "tpm": 4000000, + "output_cost_per_token_batches": 6e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query", + "supports_service_tier": true + }, + "gemini/gemini-3.1-flash-image-preview": { + "input_cost_per_token": 2.5e-07, + "input_cost_per_token_batches": 1.25e-07, + "litellm_provider": "gemini", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.045, + "output_cost_per_image_token": 6e-05, + "output_cost_per_image_token_batches": 3e-05, + "output_cost_per_token": 1.5e-06, + "output_cost_per_token_batches": 7.5e-07, + "rpm": 1000, + "tpm": 4000000, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-3.1-flash-image-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini/deep-research-pro-preview-12-2025": { + "input_cost_per_image": 0.0011, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "gemini", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.134, + "output_cost_per_image_token": 0.00012, + "output_cost_per_token": 1.2e-05, + "rpm": 1000, + "tpm": 4000000, + "output_cost_per_token_batches": 6e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_web_search": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.5-flash-lite": { + "cache_read_input_token_cost": 1e-08, + "input_cost_per_audio_token": 3e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "rpm": 15, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-lite", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + }, + "supports_service_tier": true + }, + "gemini/gemini-2.5-flash-lite-preview-09-2025": { + "cache_read_input_token_cost": 1e-08, + "input_cost_per_audio_token": 3e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "rpm": 15, + "source": "https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.5-flash-preview-09-2025": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 15, + "source": "https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-flash-latest": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 15, + "source": "https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-flash-lite-latest": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_audio_token": 3e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "rpm": 15, + "source": "https://developers.googleblog.com/en/continuing-to-bring-you-our-latest-models-with-an-improved-gemini-2-5-flash-and-flash-lite-release/", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.5-flash-lite-preview-06-17": { + "deprecation_date": "2025-11-18", + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_audio_token": 5e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "rpm": 15, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-lite", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.5-flash-preview-tts": { + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "mode": "audio_speech", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/audio/speech" + ], + "tpm": 4000000, + "rpm": 10 + }, + "gemini/gemini-2.5-pro": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "input_cost_per_token_priority": 1.25e-06, + "input_cost_per_token_above_200k_tokens_priority": 2.5e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "output_cost_per_token_priority": 1e-05, + "output_cost_per_token_above_200k_tokens_priority": 1.5e-05, + "rpm": 2000, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supports_service_tier": true, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 800000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-2.5-computer-use-preview-10-2025": { + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "gemini", + "max_images_per_prompt": 3000, + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "rpm": 2000, + "source": "https://ai.google.dev/gemini-api/docs/computer-use", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_computer_use": true, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "tpm": 800000 + }, + "gemini/gemini-3-pro-preview": { + "deprecation_date": "2026-03-09", + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "rpm": 2000, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 800000, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini/gemini-3.1-flash-lite-preview": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_per_audio_token": 5e-08, + "input_cost_per_audio_token": 5e-07, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 1.5e-06, + "output_cost_per_token": 1.5e-06, + "rpm": 15, + "source": "https://ai.google.dev/gemini-api/docs/models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_audio_output": false, + "supports_code_execution": true, + "supports_file_search": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query", + "supports_service_tier": true + }, + "gemini/gemini-3-flash-preview": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 5e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 3e-06, + "output_cost_per_token": 3e-06, + "rpm": 2000, + "source": "https://ai.google.dev/pricing/gemini-3", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "tpm": 800000, + "input_cost_per_token_priority": 9e-07, + "input_cost_per_audio_token_priority": 1.8e-06, + "output_cost_per_token_priority": 5.4e-06, + "cache_read_input_token_cost_priority": 9e-08, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini/gemini-3.1-pro-preview": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "rpm": 2000, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-3.1-pro-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_url_context": true, + "supports_native_streaming": true, + "tpm": 800000, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini/gemini-3.1-pro-preview-customtools": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "rpm": 2000, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-3.1-pro-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_url_context": true, + "supports_native_streaming": true, + "tpm": 800000, + "input_cost_per_token_priority": 3.6e-06, + "input_cost_per_token_above_200k_tokens_priority": 7.2e-06, + "output_cost_per_token_priority": 2.16e-05, + "output_cost_per_token_above_200k_tokens_priority": 3.24e-05, + "cache_read_input_token_cost_priority": 3.6e-07, + "cache_read_input_token_cost_above_200k_tokens_priority": 7.2e-07, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini-3-flash-preview": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 3e-06, + "output_cost_per_token": 3e-06, + "source": "https://ai.google.dev/pricing/gemini-3", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "input_cost_per_token_priority": 9e-07, + "input_cost_per_audio_token_priority": 1.8e-06, + "output_cost_per_token_priority": 5.4e-06, + "cache_read_input_token_cost_priority": 9e-08, + "supports_service_tier": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "gemini/gemini-2.5-pro-preview-tts": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "rpm": 10000, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-2.5-pro-preview", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "audio" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 10000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-exp-1114": { + "input_cost_per_token": 0, + "input_cost_per_token_above_128k_tokens": 0, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "metadata": { + "notes": "Rate limits not documented for gemini-exp-1114. Assuming same as gemini-1.5-pro.", + "supports_tool_choice": true + }, + "mode": "chat", + "output_cost_per_token": 0, + "output_cost_per_token_above_128k_tokens": 0, + "rpm": 1000, + "source": "https://ai.google.dev/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "tpm": 4000000 + }, + "gemini/gemini-exp-1206": { + "input_cost_per_token": 0, + "input_cost_per_token_above_128k_tokens": 0, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 2097152, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "metadata": { + "notes": "Rate limits not documented for gemini-exp-1206. Assuming same as gemini-1.5-pro.", + "supports_tool_choice": true + }, + "mode": "chat", + "output_cost_per_token": 0, + "output_cost_per_token_above_128k_tokens": 0, + "rpm": 1000, + "source": "https://ai.google.dev/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "tpm": 4000000 + }, + "gemini/gemini-gemma-2-27b-it": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "gemini", + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.05e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini/gemini-gemma-2-9b-it": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "gemini", + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.05e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini/gemma-3-27b-it": { + "input_cost_per_audio_per_second": 0, + "input_cost_per_audio_per_second_above_128k_tokens": 0, + "input_cost_per_character": 0, + "input_cost_per_character_above_128k_tokens": 0, + "input_cost_per_image": 0, + "input_cost_per_image_above_128k_tokens": 0, + "input_cost_per_token": 0, + "input_cost_per_token_above_128k_tokens": 0, + "input_cost_per_video_per_second": 0, + "input_cost_per_video_per_second_above_128k_tokens": 0, + "litellm_provider": "gemini", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_character": 0, + "output_cost_per_character_above_128k_tokens": 0, + "output_cost_per_token": 0, + "output_cost_per_token_above_128k_tokens": 0, + "source": "https://aistudio.google.com", + "supports_audio_output": false, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "gemini/imagen-3.0-fast-generate-001": { + "litellm_provider": "gemini", + "mode": "image_generation", + "output_cost_per_image": 0.02, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "gemini/imagen-3.0-generate-001": { + "litellm_provider": "gemini", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "gemini/imagen-3.0-generate-002": { + "deprecation_date": "2025-11-10", + "litellm_provider": "gemini", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "gemini/imagen-4.0-fast-generate-001": { + "litellm_provider": "gemini", + "mode": "image_generation", + "output_cost_per_image": 0.02, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "gemini/imagen-4.0-generate-001": { + "litellm_provider": "gemini", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "gemini/imagen-4.0-ultra-generate-001": { + "litellm_provider": "gemini", + "mode": "image_generation", + "output_cost_per_image": 0.06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "gemini/learnlm-1.5-pro-experimental": { + "input_cost_per_audio_per_second": 0, + "input_cost_per_audio_per_second_above_128k_tokens": 0, + "input_cost_per_character": 0, + "input_cost_per_character_above_128k_tokens": 0, + "input_cost_per_image": 0, + "input_cost_per_image_above_128k_tokens": 0, + "input_cost_per_token": 0, + "input_cost_per_token_above_128k_tokens": 0, + "input_cost_per_video_per_second": 0, + "input_cost_per_video_per_second_above_128k_tokens": 0, + "litellm_provider": "gemini", + "max_input_tokens": 32767, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_character": 0, + "output_cost_per_character_above_128k_tokens": 0, + "output_cost_per_token": 0, + "output_cost_per_token_above_128k_tokens": 0, + "source": "https://aistudio.google.com", + "supports_audio_output": false, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gemini/lyria-3-clip-preview": { + "input_cost_per_token": 0, + "litellm_provider": "gemini", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_image": 0.04, + "output_cost_per_token": 0, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "audio" + ], + "supports_audio_input": false, + "supports_audio_output": true, + "supports_function_calling": false, + "supports_prompt_caching": false, + "supports_response_schema": false, + "supports_system_messages": false, + "supports_vision": false, + "supports_web_search": false + }, + "gemini/lyria-3-pro-preview": { + "input_cost_per_token": 0, + "litellm_provider": "gemini", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "audio" + ], + "supports_audio_input": false, + "supports_audio_output": true, + "supports_function_calling": false, + "supports_prompt_caching": false, + "supports_response_schema": false, + "supports_system_messages": false, + "supports_vision": false, + "supports_web_search": false + }, + "gemini/veo-2.0-generate-001": { + "litellm_provider": "gemini", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.35, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "gemini/veo-3.1-fast-generate-preview": { + "litellm_provider": "gemini", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.15, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "gemini/veo-3.1-generate-preview": { + "litellm_provider": "gemini", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.4, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "gemini/veo-3.1-lite-generate-preview": { + "litellm_provider": "gemini", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.05, + "output_cost_per_second_1080p": 0.08, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "gemini/veo-3.1-fast-generate-001": { + "litellm_provider": "gemini", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.15, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "gemini/veo-3.1-generate-001": { + "litellm_provider": "gemini", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.4, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "github_copilot/claude-haiku-4.5": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/claude-opus-4.5": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/claude-opus-4.6-fast": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/claude-opus-41": { + "litellm_provider": "github_copilot", + "max_input_tokens": 80000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_vision": true + }, + "github_copilot/claude-sonnet-4": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/claude-sonnet-4.5": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/gemini-2.5-pro": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/gemini-3-pro-preview": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/gpt-3.5-turbo": { + "litellm_provider": "github_copilot", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true + }, + "github_copilot/gpt-3.5-turbo-0613": { + "litellm_provider": "github_copilot", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true + }, + "github_copilot/gpt-4": { + "litellm_provider": "github_copilot", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true + }, + "github_copilot/gpt-4-0613": { + "litellm_provider": "github_copilot", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true + }, + "github_copilot/gpt-4-o-preview": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true + }, + "github_copilot/gpt-4.1": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-4.1-2025-04-14": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-41-copilot": { + "litellm_provider": "github_copilot", + "mode": "completion" + }, + "github_copilot/gpt-4o": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/gpt-4o-2024-05-13": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/gpt-4o-2024-08-06": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true + }, + "github_copilot/gpt-4o-2024-11-20": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "github_copilot/gpt-4o-mini": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true + }, + "github_copilot/gpt-4o-mini-2024-07-18": { + "litellm_provider": "github_copilot", + "max_input_tokens": 64000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true + }, + "github_copilot/gpt-5": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-5-mini": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-5.1": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-5.1-codex-max": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-5.2": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/gpt-5.3-codex": { + "litellm_provider": "github_copilot", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "github_copilot/text-embedding-3-small": { + "litellm_provider": "github_copilot", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding" + }, + "github_copilot/text-embedding-3-small-inference": { + "litellm_provider": "github_copilot", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding" + }, + "github_copilot/text-embedding-ada-002": { + "litellm_provider": "github_copilot", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding" + }, + "chatgpt/gpt-5.4": { + "litellm_provider": "chatgpt", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.4-pro": { + "litellm_provider": "chatgpt", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.3-codex": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.3-codex-spark": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.3-instant": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "responses", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.3-chat-latest": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "responses", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.2-codex": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.2": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "responses", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.1-codex-max": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "chatgpt/gpt-5.1-codex-mini": { + "litellm_provider": "chatgpt", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "responses", + "supported_endpoints": [ + "/v1/responses" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "gigachat/GigaChat-2-Lite": { + "input_cost_per_token": 0.0, + "litellm_provider": "gigachat", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true, + "supports_system_messages": true + }, + "gigachat/GigaChat-2-Max": { + "input_cost_per_token": 0.0, + "litellm_provider": "gigachat", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_vision": true + }, + "gigachat/GigaChat-2-Pro": { + "input_cost_per_token": 0.0, + "litellm_provider": "gigachat", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_vision": true + }, + "gigachat/Embeddings": { + "input_cost_per_token": 0.0, + "litellm_provider": "gigachat", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024 + }, + "gigachat/Embeddings-2": { + "input_cost_per_token": 0.0, + "litellm_provider": "gigachat", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024 + }, + "gigachat/EmbeddingsGigaR": { + "input_cost_per_token": 0.0, + "litellm_provider": "gigachat", + "max_input_tokens": 4096, + "max_tokens": 4096, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 2560 + }, + "gmi/anthropic/claude-opus-4.5": { + "input_cost_per_token": 5e-06, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/anthropic/claude-sonnet-4.5": { + "input_cost_per_token": 3e-06, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/anthropic/claude-sonnet-4": { + "input_cost_per_token": 3e-06, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/anthropic/claude-opus-4": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/openai/gpt-5.2": { + "input_cost_per_token": 1.75e-06, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "supports_function_calling": true + }, + "gmi/openai/gpt-5.1": { + "input_cost_per_token": 1.25e-06, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true + }, + "gmi/openai/gpt-5": { + "input_cost_per_token": 1.25e-06, + "litellm_provider": "gmi", + "max_input_tokens": 409600, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true + }, + "gmi/openai/gpt-4o": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "gmi", + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/openai/gpt-4o-mini": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "gmi", + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/deepseek-ai/DeepSeek-V3.2": { + "input_cost_per_token": 2.8e-07, + "litellm_provider": "gmi", + "max_input_tokens": 163840, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_function_calling": true + }, + "gmi/deepseek-ai/DeepSeek-V3-0324": { + "input_cost_per_token": 2.8e-07, + "litellm_provider": "gmi", + "max_input_tokens": 163840, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 8.8e-07, + "supports_function_calling": true + }, + "gmi/google/gemini-3-pro-preview": { + "input_cost_per_token": 2e-06, + "litellm_provider": "gmi", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/google/gemini-3-flash-preview": { + "input_cost_per_token": 5e-07, + "litellm_provider": "gmi", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_vision": true + }, + "gmi/moonshotai/Kimi-K2-Thinking": { + "input_cost_per_token": 8e-07, + "litellm_provider": "gmi", + "max_input_tokens": 262144, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.2e-06 + }, + "gmi/MiniMaxAI/MiniMax-M2.1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "gmi", + "max_input_tokens": 196608, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.2e-06 + }, + "baseten/MiniMaxAI/MiniMax-M2.5": { + "input_cost_per_token": 3e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 1.2e-06 + }, + "baseten/nvidia/Nemotron-120B-A12B": { + "input_cost_per_token": 3e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 7.5e-07 + }, + "baseten/zai-org/GLM-5": { + "input_cost_per_token": 9.5e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 3.15e-06 + }, + "baseten/zai-org/GLM-4.7": { + "input_cost_per_token": 6e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 2.2e-06 + }, + "baseten/zai-org/GLM-4.6": { + "input_cost_per_token": 6e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 2.2e-06 + }, + "baseten/moonshotai/Kimi-K2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 3e-06 + }, + "baseten/moonshotai/Kimi-K2-Thinking": { + "input_cost_per_token": 6e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 2.5e-06 + }, + "baseten/moonshotai/Kimi-K2-Instruct-0905": { + "input_cost_per_token": 6e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 2.5e-06 + }, + "baseten/openai/gpt-oss-120b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 5e-07 + }, + "baseten/deepseek-ai/DeepSeek-V3.1": { + "input_cost_per_token": 5e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 1.5e-06 + }, + "baseten/deepseek-ai/DeepSeek-V3-0324": { + "input_cost_per_token": 7.7e-07, + "litellm_provider": "baseten", + "mode": "chat", + "output_cost_per_token": 7.7e-07 + }, + "gmi/Qwen/Qwen3-VL-235B-A22B-Instruct-FP8": { + "input_cost_per_token": 3e-07, + "litellm_provider": "gmi", + "max_input_tokens": 262144, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-06, + "supports_vision": true + }, + "gmi/zai-org/GLM-4.7-FP8": { + "input_cost_per_token": 4e-07, + "litellm_provider": "gmi", + "max_input_tokens": 202752, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2e-06 + }, + "google.gemma-3-12b-it": { + "input_cost_per_token": 9e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.9e-07, + "supports_system_messages": true, + "supports_vision": true + }, + "google.gemma-3-27b-it": { + "input_cost_per_token": 2.3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.8e-07, + "supports_system_messages": true, + "supports_vision": true + }, + "google.gemma-3-4b-it": { + "input_cost_per_token": 4e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 8e-08, + "supports_system_messages": true, + "supports_vision": true + }, + "google_pse/search": { + "input_cost_per_query": 0.005, + "litellm_provider": "google_pse", + "mode": "search" + }, + "global.anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_1hr": 6e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.2e-05, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "global.anthropic.claude-sonnet-4-20250514-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "global.anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_creation_input_token_cost_above_1hr": 2e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "global.amazon.nova-2-lite-v1:0": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_video_input": true, + "supports_vision": true + }, + "gpt-3.5-turbo": { + "input_cost_per_token": 5e-07, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-3.5-turbo-0125": { + "input_cost_per_token": 5e-07, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-3.5-turbo-1106": { + "deprecation_date": "2026-09-28", + "input_cost_per_token": 1e-06, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-3.5-turbo-16k": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openai", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-3.5-turbo-instruct": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "text-completion-openai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "gpt-3.5-turbo-instruct-0914": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "text-completion-openai", + "max_input_tokens": 8192, + "max_output_tokens": 4097, + "max_tokens": 4097, + "mode": "completion", + "output_cost_per_token": 2e-06 + }, + "gpt-4": { + "input_cost_per_token": 3e-05, + "litellm_provider": "openai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4-0125-preview": { + "deprecation_date": "2026-03-26", + "input_cost_per_token": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4-0314": { + "deprecation_date": "2026-03-26", + "input_cost_per_token": 3e-05, + "litellm_provider": "openai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4-0613": { + "deprecation_date": "2025-06-06", + "input_cost_per_token": 3e-05, + "litellm_provider": "openai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4-1106-preview": { + "deprecation_date": "2026-03-26", + "input_cost_per_token": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4-turbo": { + "input_cost_per_token": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gpt-4-turbo-2024-04-09": { + "input_cost_per_token": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gpt-4-turbo-preview": { + "input_cost_per_token": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4.1": { + "cache_read_input_token_cost": 5e-07, + "cache_read_input_token_cost_priority": 8.75e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-06, + "output_cost_per_token_batches": 4e-06, + "output_cost_per_token_priority": 1.4e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "gpt-4.1-2025-04-14": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-06, + "output_cost_per_token_batches": 4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "gpt-4.1-mini": { + "cache_read_input_token_cost": 1e-07, + "cache_read_input_token_cost_priority": 1.75e-07, + "input_cost_per_token": 4e-07, + "input_cost_per_token_batches": 2e-07, + "input_cost_per_token_priority": 7e-07, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "output_cost_per_token_batches": 8e-07, + "output_cost_per_token_priority": 2.8e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "gpt-4.1-mini-2025-04-14": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 4e-07, + "input_cost_per_token_batches": 2e-07, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "output_cost_per_token_batches": 8e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "gpt-4.1-nano": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_priority": 5e-08, + "input_cost_per_token": 1e-07, + "input_cost_per_token_batches": 5e-08, + "input_cost_per_token_priority": 2e-07, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "output_cost_per_token_batches": 2e-07, + "output_cost_per_token_priority": 8e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4.1-nano-2025-04-14": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 1e-07, + "input_cost_per_token_batches": 5e-08, + "litellm_provider": "openai", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "output_cost_per_token_batches": 2e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4o": { + "cache_read_input_token_cost": 1.25e-06, + "cache_read_input_token_cost_priority": 2.125e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_batches": 1.25e-06, + "input_cost_per_token_priority": 4.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_batches": 5e-06, + "output_cost_per_token_priority": 1.7e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4o-2024-05-13": { + "input_cost_per_token": 5e-06, + "input_cost_per_token_batches": 2.5e-06, + "input_cost_per_token_priority": 8.75e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_batches": 7.5e-06, + "output_cost_per_token_priority": 2.625e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gpt-4o-2024-08-06": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_batches": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_batches": 5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4o-2024-11-20": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_batches": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_batches": 5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4o-audio-preview": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-audio-preview-2024-12-17": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-audio-preview-2025-06-03": { + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 1e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-audio": { + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses", + "/v1/realtime", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "gpt-audio-1.5": { + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "gpt-audio-2025-08-28": { + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses", + "/v1/realtime", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "gpt-audio-mini": { + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses", + "/v1/realtime", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "gpt-audio-mini-2025-10-06": { + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses", + "/v1/realtime", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "gpt-audio-mini-2025-12-15": { + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses", + "/v1/realtime", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": false, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "gpt-4o-mini": { + "cache_read_input_token_cost": 7.5e-08, + "cache_read_input_token_cost_priority": 1.25e-07, + "input_cost_per_token": 1.5e-07, + "input_cost_per_token_batches": 7.5e-08, + "input_cost_per_token_priority": 2.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "output_cost_per_token_batches": 3e-07, + "output_cost_per_token_priority": 1e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4o-mini-2024-07-18": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.5e-07, + "input_cost_per_token_batches": 7.5e-08, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "output_cost_per_token_batches": 3e-07, + "search_context_cost_per_query": { + "search_context_size_high": 0.03, + "search_context_size_low": 0.025, + "search_context_size_medium": 0.0275 + }, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-4o-mini-audio-preview": { + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 6e-07, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-mini-audio-preview-2024-12-17": { + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 6e-07, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-mini-realtime-preview": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-mini-realtime-preview-2024-12-17": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-mini-search-preview": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.5e-07, + "input_cost_per_token_batches": 7.5e-08, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "output_cost_per_token_batches": 3e-07, + "search_context_cost_per_query": { + "search_context_size_high": 0.03, + "search_context_size_low": 0.025, + "search_context_size_medium": 0.0275 + }, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "gpt-4o-mini-search-preview-2025-03-11": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.5e-07, + "input_cost_per_token_batches": 7.5e-08, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "output_cost_per_token_batches": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gpt-4o-mini-transcribe": { + "input_cost_per_audio_token": 3e-06, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 5e-06, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "gpt-4o-mini-tts": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "mode": "audio_speech", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_second": 0.00025, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/speech" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "audio" + ] + }, + "gpt-4o-realtime-preview": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 2e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-realtime-preview-2024-12-17": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 2e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-realtime-preview-2025-06-03": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_audio_token": 4e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 8e-05, + "output_cost_per_token": 2e-05, + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-4o-search-preview": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_batches": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_batches": 5e-06, + "search_context_cost_per_query": { + "search_context_size_high": 0.05, + "search_context_size_low": 0.03, + "search_context_size_medium": 0.035 + }, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "gpt-4o-search-preview-2025-03-11": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_batches": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_batches": 5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "gpt-4o-transcribe": { + "input_cost_per_audio_token": 6e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "gpt-image-1.5": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_token": 1e-05, + "input_cost_per_image_token": 8e-06, + "output_cost_per_image_token": 3.2e-05, + "supported_endpoints": [ + "/v1/images/generations" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "gpt-image-1.5-2025-12-16": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_token": 1e-05, + "input_cost_per_image_token": 8e-06, + "output_cost_per_image_token": 3.2e-05, + "supported_endpoints": [ + "/v1/images/generations" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "gpt-image-2": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_token": 1e-05, + "input_cost_per_image_token": 8e-06, + "output_cost_per_image_token": 3e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "gpt-image-2-2026-04-21": { + "cache_read_input_image_token_cost": 2e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_token": 1e-05, + "input_cost_per_image_token": 8e-06, + "output_cost_per_image_token": 3e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "low/1024-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.009, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "low/1024-x-1536/gpt-image-1.5": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "low/1536-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "medium/1024-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.034, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "medium/1024-x-1536/gpt-image-1.5": { + "input_cost_per_image": 0.05, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "medium/1536-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.05, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "high/1024-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.133, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "high/1024-x-1536/gpt-image-1.5": { + "input_cost_per_image": 0.2, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "high/1536-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.2, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "standard/1024-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.009, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "standard/1024-x-1536/gpt-image-1.5": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "standard/1536-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "1024-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.009, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "1024-x-1536/gpt-image-1.5": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "1536-x-1024/gpt-image-1.5": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "low/1024-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.009, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "low/1024-x-1536/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "low/1536-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "medium/1024-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.034, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "medium/1024-x-1536/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.05, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "medium/1536-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.05, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "high/1024-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.133, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "high/1024-x-1536/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.2, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "high/1536-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.2, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "standard/1024-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.009, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "standard/1024-x-1536/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "standard/1536-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "1024-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.009, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "1024-x-1536/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "1536-x-1024/gpt-image-1.5-2025-12-16": { + "input_cost_per_image": 0.013, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ], + "supports_vision": true, + "supports_pdf_input": true + }, + "gpt-5": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_flex": 6.25e-08, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_flex": 6.25e-07, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_flex": 5e-06, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.1": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.1-2025-11-13": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.1-chat-latest": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": false, + "supports_native_streaming": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": false, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.2": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.2-2025-12-11": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.2-chat-latest": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.3-chat-latest": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.2-pro": { + "input_cost_per_token": 2.1e-05, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.000168, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.2-pro-2025-12-11": { + "input_cost_per_token": 2.1e-05, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.000168, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.5": { + "cache_read_input_token_cost": 5e-07, + "cache_read_input_token_cost_above_272k_tokens": 1e-06, + "cache_read_input_token_cost_flex": 2.5e-07, + "cache_read_input_token_cost_priority": 1e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_token_above_272k_tokens": 1e-05, + "input_cost_per_token_flex": 2.5e-06, + "input_cost_per_token_batches": 2.5e-06, + "input_cost_per_token_priority": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-05, + "output_cost_per_token_above_272k_tokens": 4.5e-05, + "output_cost_per_token_flex": 1.5e-05, + "output_cost_per_token_batches": 1.5e-05, + "output_cost_per_token_priority": 6e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.5-2026-04-23": { + "cache_read_input_token_cost": 5e-07, + "cache_read_input_token_cost_above_272k_tokens": 1e-06, + "cache_read_input_token_cost_flex": 2.5e-07, + "cache_read_input_token_cost_priority": 1e-06, + "input_cost_per_token": 5e-06, + "input_cost_per_token_above_272k_tokens": 1e-05, + "input_cost_per_token_flex": 2.5e-06, + "input_cost_per_token_batches": 2.5e-06, + "input_cost_per_token_priority": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-05, + "output_cost_per_token_above_272k_tokens": 4.5e-05, + "output_cost_per_token_flex": 1.5e-05, + "output_cost_per_token_batches": 1.5e-05, + "output_cost_per_token_priority": 6e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.5-pro": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "input_cost_per_token_flex": 1.5e-05, + "input_cost_per_token_batches": 1.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "output_cost_per_token_flex": 9e-05, + "output_cost_per_token_batches": 9e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.5-pro-2026-04-23": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "input_cost_per_token_flex": 1.5e-05, + "input_cost_per_token_batches": 1.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "output_cost_per_token_flex": 9e-05, + "output_cost_per_token_batches": 9e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.4": { + "cache_read_input_token_cost": 2.5e-07, + "cache_read_input_token_cost_above_272k_tokens": 5e-07, + "cache_read_input_token_cost_flex": 1.3e-07, + "cache_read_input_token_cost_priority": 5e-07, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_above_272k_tokens": 5e-06, + "input_cost_per_token_flex": 1.25e-06, + "input_cost_per_token_batches": 1.25e-06, + "input_cost_per_token_priority": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_272k_tokens": 2.25e-05, + "output_cost_per_token_flex": 7.5e-06, + "output_cost_per_token_batches": 7.5e-06, + "output_cost_per_token_priority": 3e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.4-2026-03-05": { + "cache_read_input_token_cost": 2.5e-07, + "cache_read_input_token_cost_above_272k_tokens": 5e-07, + "cache_read_input_token_cost_flex": 1.3e-07, + "cache_read_input_token_cost_priority": 5e-07, + "input_cost_per_token": 2.5e-06, + "input_cost_per_token_above_272k_tokens": 5e-06, + "input_cost_per_token_flex": 1.25e-06, + "input_cost_per_token_batches": 1.25e-06, + "input_cost_per_token_priority": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_272k_tokens": 2.25e-05, + "output_cost_per_token_flex": 7.5e-06, + "output_cost_per_token_batches": 7.5e-06, + "output_cost_per_token_priority": 3e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true + }, + "gpt-5.4-pro": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "input_cost_per_token_flex": 1.5e-05, + "input_cost_per_token_batches": 1.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "output_cost_per_token_flex": 9e-05, + "output_cost_per_token_batches": 9e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.4-pro-2026-03-05": { + "cache_read_input_token_cost": 3e-06, + "cache_read_input_token_cost_above_272k_tokens": 6e-06, + "input_cost_per_token": 3e-05, + "input_cost_per_token_above_272k_tokens": 6e-05, + "input_cost_per_token_flex": 1.5e-05, + "input_cost_per_token_batches": 1.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 1050000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 0.00018, + "output_cost_per_token_above_272k_tokens": 0.00027, + "output_cost_per_token_flex": 9e-05, + "output_cost_per_token_batches": 9e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.4-mini": { + "cache_read_input_token_cost": 7.5e-08, + "cache_read_input_token_cost_flex": 3.75e-08, + "cache_read_input_token_cost_batches": 3.75e-08, + "cache_read_input_token_cost_priority": 1.5e-07, + "input_cost_per_token": 7.5e-07, + "input_cost_per_token_flex": 3.75e-07, + "input_cost_per_token_batches": 3.75e-07, + "input_cost_per_token_priority": 1.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4.5e-06, + "output_cost_per_token_flex": 2.25e-06, + "output_cost_per_token_batches": 2.25e-06, + "output_cost_per_token_priority": 9e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": false + }, + "gpt-5.4-mini-2026-03-17": { + "cache_read_input_token_cost": 7.5e-08, + "cache_read_input_token_cost_flex": 3.75e-08, + "cache_read_input_token_cost_batches": 3.75e-08, + "cache_read_input_token_cost_priority": 1.5e-07, + "input_cost_per_token": 7.5e-07, + "input_cost_per_token_flex": 3.75e-07, + "input_cost_per_token_batches": 3.75e-07, + "input_cost_per_token_priority": 1.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4.5e-06, + "output_cost_per_token_flex": 2.25e-06, + "output_cost_per_token_batches": 2.25e-06, + "output_cost_per_token_priority": 9e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": false + }, + "gpt-5.4-nano": { + "cache_read_input_token_cost": 2e-08, + "cache_read_input_token_cost_flex": 1e-08, + "cache_read_input_token_cost_batches": 1e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_flex": 1e-07, + "input_cost_per_token_batches": 1e-07, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "output_cost_per_token_flex": 6.25e-07, + "output_cost_per_token_batches": 6.25e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": false + }, + "gpt-5.4-nano-2026-03-17": { + "cache_read_input_token_cost": 2e-08, + "cache_read_input_token_cost_flex": 1e-08, + "cache_read_input_token_cost_batches": 1e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_flex": 1e-07, + "input_cost_per_token_batches": 1e-07, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "output_cost_per_token_flex": 6.25e-07, + "output_cost_per_token_batches": 6.25e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": true, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": false + }, + "gpt-5-pro": { + "input_cost_per_token": 1.5e-05, + "input_cost_per_token_batches": 7.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 272000, + "max_tokens": 272000, + "mode": "responses", + "output_cost_per_token": 0.00012, + "output_cost_per_token_batches": 6e-05, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": false, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-pro-2025-10-06": { + "input_cost_per_token": 1.5e-05, + "input_cost_per_token_batches": 7.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 272000, + "max_tokens": 272000, + "mode": "responses", + "output_cost_per_token": 0.00012, + "output_cost_per_token_batches": 6e-05, + "supported_endpoints": [ + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": false, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-2025-08-07": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_flex": 6.25e-08, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_flex": 6.25e-07, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_flex": 5e-06, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-chat": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": false, + "supports_native_streaming": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": false, + "supports_vision": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-chat-latest": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": false, + "supports_native_streaming": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": false, + "supports_vision": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-codex": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.1-codex": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_priority": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "output_cost_per_token_priority": 2e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.1-codex-max": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.1-codex-mini": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_priority": 4.5e-08, + "input_cost_per_token": 2.5e-07, + "input_cost_per_token_priority": 4.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 2e-06, + "output_cost_per_token_priority": 3.6e-06, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.2-codex": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "gpt-5.3-codex": { + "cache_read_input_token_cost": 1.75e-07, + "cache_read_input_token_cost_priority": 3.5e-07, + "input_cost_per_token": 1.75e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "responses", + "output_cost_per_token": 1.4e-05, + "output_cost_per_token_priority": 2.8e-05, + "supported_endpoints": [ + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": false, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-mini": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_flex": 1.25e-08, + "cache_read_input_token_cost_priority": 4.5e-08, + "input_cost_per_token": 2.5e-07, + "input_cost_per_token_flex": 1.25e-07, + "input_cost_per_token_priority": 4.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "output_cost_per_token_flex": 1e-06, + "output_cost_per_token_priority": 3.6e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-mini-2025-08-07": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_flex": 1.25e-08, + "cache_read_input_token_cost_priority": 4.5e-08, + "input_cost_per_token": 2.5e-07, + "input_cost_per_token_flex": 1.25e-07, + "input_cost_per_token_priority": 4.5e-07, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "output_cost_per_token_flex": 1e-06, + "output_cost_per_token_priority": 3.6e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-nano": { + "cache_read_input_token_cost": 5e-09, + "cache_read_input_token_cost_flex": 2.5e-09, + "input_cost_per_token": 5e-08, + "input_cost_per_token_flex": 2.5e-08, + "input_cost_per_token_priority": 2.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "output_cost_per_token_flex": 2e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-nano-2025-08-07": { + "cache_read_input_token_cost": 5e-09, + "cache_read_input_token_cost_flex": 2.5e-09, + "input_cost_per_token": 5e-08, + "input_cost_per_token_flex": 2.5e-08, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "output_cost_per_token_flex": 2e-07, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-image-1": { + "cache_read_input_image_token_cost": 2.5e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_image_token": 1e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_image_token": 4e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "gpt-image-1-mini": { + "cache_read_input_image_token_cost": 2.5e-07, + "cache_read_input_token_cost": 2e-07, + "input_cost_per_image_token": 2.5e-06, + "input_cost_per_token": 2e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_image_token": 8e-06, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "gpt-realtime": { + "cache_creation_input_audio_token_cost": 4e-07, + "cache_read_input_token_cost": 4e-07, + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_image": 5e-06, + "input_cost_per_token": 4e-06, + "litellm_provider": "openai", + "max_input_tokens": 32000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1.6e-05, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-realtime-1.5": { + "cache_creation_input_audio_token_cost": 4e-07, + "cache_read_input_token_cost": 4e-07, + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_image": 5e-06, + "input_cost_per_token": 4e-06, + "litellm_provider": "openai", + "max_input_tokens": 32000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1.6e-05, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-realtime-mini": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_audio_token_cost": 3e-07, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-realtime-2025-08-28": { + "cache_creation_input_audio_token_cost": 4e-07, + "cache_read_input_token_cost": 4e-07, + "input_cost_per_audio_token": 3.2e-05, + "input_cost_per_image": 5e-06, + "input_cost_per_token": 4e-06, + "litellm_provider": "openai", + "max_input_tokens": 32000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 6.4e-05, + "output_cost_per_token": 1.6e-05, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gradient_ai/alibaba-qwen3-32b": { + "litellm_provider": "gradient_ai", + "max_tokens": 40960, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 131072, + "max_output_tokens": 40960 + }, + "gradient_ai/anthropic-claude-3-opus": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "gradient_ai", + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 200000, + "max_output_tokens": 1024 + }, + "gradient_ai/anthropic-claude-3.5-haiku": { + "input_cost_per_token": 8e-07, + "litellm_provider": "gradient_ai", + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 200000, + "max_output_tokens": 1024 + }, + "gradient_ai/anthropic-claude-3.5-sonnet": { + "input_cost_per_token": 3e-06, + "litellm_provider": "gradient_ai", + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 200000, + "max_output_tokens": 1024 + }, + "gradient_ai/anthropic-claude-3.7-sonnet": { + "input_cost_per_token": 3e-06, + "litellm_provider": "gradient_ai", + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 200000, + "max_output_tokens": 1024 + }, + "gradient_ai/deepseek-r1-distill-llama-70b": { + "input_cost_per_token": 9.9e-07, + "litellm_provider": "gradient_ai", + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 9.9e-07, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 32768, + "max_output_tokens": 8000 + }, + "gradient_ai/llama3-8b-instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "gradient_ai", + "max_tokens": 512, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 8192, + "max_output_tokens": 512 + }, + "gradient_ai/llama3.3-70b-instruct": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "gradient_ai", + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 6.5e-07, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 128000, + "max_output_tokens": 2048 + }, + "gradient_ai/mistral-nemo-instruct-2407": { + "input_cost_per_token": 3e-07, + "litellm_provider": "gradient_ai", + "max_tokens": 512, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 128000, + "max_output_tokens": 512 + }, + "gradient_ai/openai-gpt-4o": { + "litellm_provider": "gradient_ai", + "max_tokens": 16384, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 128000, + "max_output_tokens": 16384 + }, + "gradient_ai/openai-gpt-4o-mini": { + "litellm_provider": "gradient_ai", + "max_tokens": 16384, + "mode": "chat", + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 128000, + "max_output_tokens": 16384 + }, + "gradient_ai/openai-o3": { + "input_cost_per_token": 2e-06, + "litellm_provider": "gradient_ai", + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 200000, + "max_output_tokens": 100000 + }, + "gradient_ai/openai-o3-mini": { + "input_cost_per_token": 1.1e-06, + "litellm_provider": "gradient_ai", + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supported_endpoints": [ + "/v1/chat/completions" + ], + "supported_modalities": [ + "text" + ], + "supports_tool_choice": false, + "max_input_tokens": 200000, + "max_output_tokens": 100000 + }, + "lemonade/Qwen3-Coder-30B-A3B-Instruct-GGUF": { + "input_cost_per_token": 0, + "litellm_provider": "lemonade", + "max_tokens": 32768, + "max_input_tokens": 262144, + "max_output_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "lemonade/gpt-oss-20b-mxfp4-GGUF": { + "input_cost_per_token": 0, + "litellm_provider": "lemonade", + "max_tokens": 32768, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "lemonade/gpt-oss-120b-mxfp-GGUF": { + "input_cost_per_token": 0, + "litellm_provider": "lemonade", + "max_tokens": 32768, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "lemonade/Gemma-3-4b-it-GGUF": { + "input_cost_per_token": 0, + "litellm_provider": "lemonade", + "max_tokens": 8192, + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "lemonade/Qwen3-4B-Instruct-2507-GGUF": { + "input_cost_per_token": 0, + "litellm_provider": "lemonade", + "max_tokens": 32768, + "max_input_tokens": 262144, + "max_output_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "amazon-nova/nova-micro-v1": { + "input_cost_per_token": 3.5e-08, + "litellm_provider": "amazon_nova", + "max_input_tokens": 128000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.4e-07, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true + }, + "amazon-nova/nova-lite-v1": { + "input_cost_per_token": 6e-08, + "litellm_provider": "amazon_nova", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 2.4e-07, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "amazon-nova/nova-premier-v1": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "amazon_nova", + "max_input_tokens": 1000000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.25e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": false, + "supports_response_schema": true, + "supports_vision": true + }, + "amazon-nova/nova-pro-v1": { + "input_cost_per_token": 8e-07, + "litellm_provider": "amazon_nova", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "groq/llama-3.1-8b-instant": { + "input_cost_per_token": 5e-08, + "litellm_provider": "groq", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 8e-08, + "supports_function_calling": true, + "supports_response_schema": false, + "supports_tool_choice": true + }, + "groq/llama-3.3-70b-versatile": { + "input_cost_per_token": 5.9e-07, + "litellm_provider": "groq", + "max_input_tokens": 128000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 7.9e-07, + "supports_function_calling": true, + "supports_response_schema": false, + "supports_tool_choice": true + }, + "groq/gemma-7b-it": { + "input_cost_per_token": 5e-08, + "litellm_provider": "groq", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 8e-08, + "supports_function_calling": true, + "supports_response_schema": false, + "supports_tool_choice": true + }, + "groq/meta-llama/llama-guard-4-12b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "groq", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-07 + }, + "groq/meta-llama/llama-4-maverick-17b-128e-instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "groq", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "groq/meta-llama/llama-4-scout-17b-16e-instruct": { + "input_cost_per_token": 1.1e-07, + "litellm_provider": "groq", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.4e-07, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "groq/moonshotai/kimi-k2-instruct-0905": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3e-06, + "cache_read_input_token_cost": 5e-07, + "litellm_provider": "groq", + "max_input_tokens": 262144, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "groq/openai/gpt-oss-120b": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "groq", + "max_input_tokens": 131072, + "max_output_tokens": 32766, + "max_tokens": 32766, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "groq/openai/gpt-oss-20b": { + "cache_read_input_token_cost": 3.75e-08, + "input_cost_per_token": 7.5e-08, + "litellm_provider": "groq", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "groq/openai/gpt-oss-safeguard-20b": { + "cache_read_input_token_cost": 3.7e-08, + "input_cost_per_token": 7.5e-08, + "litellm_provider": "groq", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "groq/playai-tts": { + "input_cost_per_character": 5e-05, + "litellm_provider": "groq", + "max_input_tokens": 10000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "audio_speech" + }, + "groq/qwen/qwen3-32b": { + "input_cost_per_token": 2.9e-07, + "litellm_provider": "groq", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 5.9e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true + }, + "groq/whisper-large-v3": { + "input_cost_per_second": 3.083e-05, + "litellm_provider": "groq", + "mode": "audio_transcription", + "output_cost_per_second": 0.0 + }, + "groq/whisper-large-v3-turbo": { + "input_cost_per_second": 1.111e-05, + "litellm_provider": "groq", + "mode": "audio_transcription", + "output_cost_per_second": 0.0 + }, + "hd/1024-x-1024/dall-e-3": { + "input_cost_per_pixel": 7.629e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "hd/1024-x-1792/dall-e-3": { + "input_cost_per_pixel": 6.539e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "hd/1792-x-1024/dall-e-3": { + "input_cost_per_pixel": 6.539e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "heroku/claude-3-5-haiku": { + "litellm_provider": "heroku", + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "max_input_tokens": 200000, + "max_output_tokens": 8192 + }, + "heroku/claude-3-5-sonnet-latest": { + "litellm_provider": "heroku", + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "max_input_tokens": 200000, + "max_output_tokens": 8192 + }, + "heroku/claude-3-7-sonnet": { + "litellm_provider": "heroku", + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "max_input_tokens": 200000, + "max_output_tokens": 8192 + }, + "heroku/claude-4-sonnet": { + "litellm_provider": "heroku", + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "max_input_tokens": 200000, + "max_output_tokens": 8192 + }, + "high/1024-x-1024/gpt-image-1": { + "input_cost_per_image": 0.167, + "input_cost_per_pixel": 1.59263611e-07, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "high/1024-x-1536/gpt-image-1": { + "input_cost_per_image": 0.25, + "input_cost_per_pixel": 1.58945719e-07, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "high/1536-x-1024/gpt-image-1": { + "input_cost_per_image": 0.25, + "input_cost_per_pixel": 1.58945719e-07, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "hyperbolic/NousResearch/Hermes-3-Llama-3.1-70B": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/Qwen/QwQ-32B": { + "input_cost_per_token": 2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/Qwen/Qwen2.5-72B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/Qwen/Qwen2.5-Coder-32B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/Qwen/Qwen3-235B-A22B": { + "input_cost_per_token": 2e-06, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/deepseek-ai/DeepSeek-R1": { + "input_cost_per_token": 4e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/deepseek-ai/DeepSeek-R1-0528": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/deepseek-ai/DeepSeek-V3": { + "input_cost_per_token": 2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/deepseek-ai/DeepSeek-V3-0324": { + "input_cost_per_token": 4e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/meta-llama/Llama-3.2-3B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/meta-llama/Llama-3.3-70B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/meta-llama/Meta-Llama-3-70B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/meta-llama/Meta-Llama-3.1-405B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/meta-llama/Meta-Llama-3.1-70B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/meta-llama/Meta-Llama-3.1-8B-Instruct": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "hyperbolic", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "hyperbolic/moonshotai/Kimi-K2-Instruct": { + "input_cost_per_token": 2e-06, + "litellm_provider": "hyperbolic", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "j2-light": { + "input_cost_per_token": 3e-06, + "litellm_provider": "ai21", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 3e-06 + }, + "j2-mid": { + "input_cost_per_token": 1e-05, + "litellm_provider": "ai21", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 1e-05 + }, + "j2-ultra": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "ai21", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 1.5e-05 + }, + "jamba-1.5": { + "input_cost_per_token": 2e-07, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "jamba-1.5-large": { + "input_cost_per_token": 2e-06, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_tool_choice": true + }, + "jamba-1.5-large@001": { + "input_cost_per_token": 2e-06, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_tool_choice": true + }, + "jamba-1.5-mini": { + "input_cost_per_token": 2e-07, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "jamba-1.5-mini@001": { + "input_cost_per_token": 2e-07, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "jamba-large-1.6": { + "input_cost_per_token": 2e-06, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_tool_choice": true + }, + "jamba-large-1.7": { + "input_cost_per_token": 2e-06, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_tool_choice": true + }, + "jamba-mini-1.6": { + "input_cost_per_token": 2e-07, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "jamba-mini-1.7": { + "input_cost_per_token": 2e-07, + "litellm_provider": "ai21", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "jina-reranker-v2-base-multilingual": { + "input_cost_per_token": 1.8e-08, + "litellm_provider": "jina_ai", + "max_document_chunks_per_query": 2048, + "max_input_tokens": 1024, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "rerank", + "output_cost_per_token": 1.8e-08 + }, + "jp.anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "input_cost_per_token_above_200k_tokens": 6.6e-06, + "output_cost_per_token_above_200k_tokens": 2.475e-05, + "cache_creation_input_token_cost_above_200k_tokens": 8.25e-06, + "cache_read_input_token_cost_above_200k_tokens": 6.6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "jp.anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.375e-06, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5.5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "lambda_ai/deepseek-llama3.3-70b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/deepseek-r1-0528": { + "input_cost_per_token": 2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/deepseek-r1-671b": { + "input_cost_per_token": 8e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 8e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/deepseek-v3-0324": { + "input_cost_per_token": 2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/hermes3-405b": { + "input_cost_per_token": 8e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 8e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/hermes3-70b": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/hermes3-8b": { + "input_cost_per_token": 2.5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-08, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/lfm-40b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/lfm-7b": { + "input_cost_per_token": 2.5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-08, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama-4-maverick-17b-128e-instruct-fp8": { + "input_cost_per_token": 5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama-4-scout-17b-16e-instruct": { + "input_cost_per_token": 5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 16384, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama3.1-405b-instruct-fp8": { + "input_cost_per_token": 8e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 8e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama3.1-70b-instruct-fp8": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama3.1-8b-instruct": { + "input_cost_per_token": 2.5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-08, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama3.1-nemotron-70b-instruct-fp8": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama3.2-11b-vision-instruct": { + "input_cost_per_token": 1.5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-08, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "lambda_ai/llama3.2-3b-instruct": { + "input_cost_per_token": 1.5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-08, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/llama3.3-70b-instruct-fp8": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/qwen25-coder-32b-instruct": { + "input_cost_per_token": 5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "lambda_ai/qwen3-32b-fp8": { + "input_cost_per_token": 5e-08, + "litellm_provider": "lambda_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "low/1024-x-1024/gpt-image-1": { + "input_cost_per_image": 0.011, + "input_cost_per_pixel": 1.0490417e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "low/1024-x-1536/gpt-image-1": { + "input_cost_per_image": 0.016, + "input_cost_per_pixel": 1.0172526e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "low/1536-x-1024/gpt-image-1": { + "input_cost_per_image": 0.016, + "input_cost_per_pixel": 1.0172526e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "max-x-max/50-steps/stability.stable-diffusion-xl-v0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.036 + }, + "max-x-max/max-steps/stability.stable-diffusion-xl-v0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.072 + }, + "medium/1024-x-1024/gpt-image-1": { + "input_cost_per_image": 0.042, + "input_cost_per_pixel": 4.0054321e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "medium/1024-x-1536/gpt-image-1": { + "input_cost_per_image": 0.063, + "input_cost_per_pixel": 4.0054321e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "medium/1536-x-1024/gpt-image-1": { + "input_cost_per_image": 0.063, + "input_cost_per_pixel": 4.0054321e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "low/1024-x-1024/gpt-image-1-mini": { + "input_cost_per_image": 0.005, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "low/1024-x-1536/gpt-image-1-mini": { + "input_cost_per_image": 0.006, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "low/1536-x-1024/gpt-image-1-mini": { + "input_cost_per_image": 0.006, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "medium/1024-x-1024/gpt-image-1-mini": { + "input_cost_per_image": 0.011, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "medium/1024-x-1536/gpt-image-1-mini": { + "input_cost_per_image": 0.015, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "medium/1536-x-1024/gpt-image-1-mini": { + "input_cost_per_image": 0.015, + "litellm_provider": "openai", + "mode": "image_generation", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "medlm-large": { + "input_cost_per_character": 5e-06, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_character": 1.5e-05, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models", + "supports_tool_choice": true + }, + "medlm-medium": { + "input_cost_per_character": 5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_character": 1e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models", + "supports_tool_choice": true + }, + "meta.llama2-13b-chat-v1": { + "input_cost_per_token": 7.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-06 + }, + "meta.llama2-70b-chat-v1": { + "input_cost_per_token": 1.95e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.56e-06 + }, + "meta.llama3-1-405b-instruct-v1:0": { + "input_cost_per_token": 5.32e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.6e-05, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama3-1-70b-instruct-v1:0": { + "input_cost_per_token": 9.9e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 9.9e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama3-1-8b-instruct-v1:0": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 2.2e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama3-2-11b-instruct-v1:0": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3.5e-07, + "supports_function_calling": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "meta.llama3-2-1b-instruct-v1:0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama3-2-3b-instruct-v1:0": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama3-2-90b-instruct-v1:0": { + "input_cost_per_token": 2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "meta.llama3-3-70b-instruct-v1:0": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama3-70b-instruct-v1:0": { + "input_cost_per_token": 2.65e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.5e-06 + }, + "meta.llama3-8b-instruct-v1:0": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07 + }, + "meta.llama4-maverick-17b-instruct-v1:0": { + "input_cost_per_token": 2.4e-07, + "input_cost_per_token_batches": 1.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 9.7e-07, + "output_cost_per_token_batches": 4.85e-07, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta.llama4-scout-17b-instruct-v1:0": { + "input_cost_per_token": 1.7e-07, + "input_cost_per_token_batches": 8.5e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6.6e-07, + "output_cost_per_token_batches": 3.3e-07, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": false + }, + "meta_llama/Llama-3.3-70B-Instruct": { + "litellm_provider": "meta_llama", + "max_input_tokens": 128000, + "max_output_tokens": 4028, + "max_tokens": 4028, + "mode": "chat", + "source": "https://llama.developer.meta.com/docs/models", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "meta_llama/Llama-3.3-8B-Instruct": { + "litellm_provider": "meta_llama", + "max_input_tokens": 128000, + "max_output_tokens": 4028, + "max_tokens": 4028, + "mode": "chat", + "source": "https://llama.developer.meta.com/docs/models", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "meta_llama/Llama-4-Maverick-17B-128E-Instruct-FP8": { + "litellm_provider": "meta_llama", + "max_input_tokens": 1000000, + "max_output_tokens": 4028, + "max_tokens": 4028, + "mode": "chat", + "source": "https://llama.developer.meta.com/docs/models", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "meta_llama/Llama-4-Scout-17B-16E-Instruct-FP8": { + "litellm_provider": "meta_llama", + "max_input_tokens": 10000000, + "max_output_tokens": 4028, + "max_tokens": 4028, + "mode": "chat", + "source": "https://llama.developer.meta.com/docs/models", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "minimax.minimax-m2": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "minimax.minimax-m2.1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 196000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "minimax/speech-02-hd": { + "input_cost_per_character": 0.0001, + "litellm_provider": "minimax", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "minimax/speech-02-turbo": { + "input_cost_per_character": 6e-05, + "litellm_provider": "minimax", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "minimax/speech-2.6-hd": { + "input_cost_per_character": 0.0001, + "litellm_provider": "minimax", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "minimax/speech-2.6-turbo": { + "input_cost_per_character": 6e-05, + "litellm_provider": "minimax", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "minimax/MiniMax-M2.1": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07, + "litellm_provider": "minimax", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_system_messages": true, + "max_input_tokens": 1000000, + "max_output_tokens": 8192 + }, + "minimax/MiniMax-M2.1-lightning": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 2.4e-06, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07, + "litellm_provider": "minimax", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_system_messages": true, + "max_input_tokens": 1000000, + "max_output_tokens": 8192 + }, + "minimax/MiniMax-M2.5": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07, + "litellm_provider": "minimax", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_system_messages": true, + "max_input_tokens": 1000000, + "max_output_tokens": 8192 + }, + "minimax/MiniMax-M2.5-lightning": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 2.4e-06, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07, + "litellm_provider": "minimax", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_system_messages": true, + "max_input_tokens": 1000000, + "max_output_tokens": 8192 + }, + "minimax/MiniMax-M2": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "cache_read_input_token_cost": 3e-08, + "cache_creation_input_token_cost": 3.75e-07, + "litellm_provider": "minimax", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_system_messages": true, + "max_input_tokens": 200000, + "max_output_tokens": 8192 + }, + "mistral.devstral-2-123b": { + "input_cost_per_token": 4e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 256000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "mistral.magistral-small-2509": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true + }, + "mistral.ministral-3-14b-instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "mistral.ministral-3-3b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "mistral.ministral-3-8b-instruct": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "mistral.mistral-7b-instruct-v0:2": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_tool_choice": true + }, + "mistral.mistral-large-2402-v1:0": { + "input_cost_per_token": 8e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_function_calling": true + }, + "mistral.mistral-large-2407-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 9e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "mistral.mistral-large-3-675b-instruct": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "mistral.mistral-small-2402-v1:0": { + "input_cost_per_token": 1e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true + }, + "mistral.mixtral-8x7b-instruct-v0:1": { + "input_cost_per_token": 4.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 7e-07, + "supports_tool_choice": true + }, + "mistral.voxtral-mini-3b-2507": { + "input_cost_per_token": 4e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4e-08, + "supports_audio_input": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "mistral.voxtral-small-24b-2507": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_audio_input": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "mistral/codestral-2405": { + "input_cost_per_token": 1e-06, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/codestral-2508": { + "input_cost_per_token": 3e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 9e-07, + "source": "https://mistral.ai/news/codestral-25-08", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/codestral-latest": { + "input_cost_per_token": 1e-06, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/codestral-mamba-latest": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "source": "https://mistral.ai/technology/", + "supports_assistant_prefill": true, + "supports_tool_choice": true + }, + "mistral/devstral-medium-2507": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://mistral.ai/news/devstral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/devstral-small-2505": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://mistral.ai/news/devstral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/devstral-small-2507": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://mistral.ai/news/devstral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/devstral-small-latest": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://docs.mistral.ai/models/devstral-small-2-25-12", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/labs-devstral-small-2512": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://docs.mistral.ai/models/devstral-small-2-25-12", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/devstral-latest": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://mistral.ai/news/devstral-2-vibe-cli", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/devstral-medium-latest": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://mistral.ai/news/devstral-2-vibe-cli", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/devstral-2512": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://mistral.ai/news/devstral-2-vibe-cli", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/magistral-medium-2506": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://mistral.ai/news/magistral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/magistral-medium-2509": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://mistral.ai/news/magistral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/magistral-medium-1-2-2509": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://mistral.ai/news/magistral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-ocr-latest": { + "litellm_provider": "mistral", + "ocr_cost_per_page": 0.001, + "annotation_cost_per_page": 0.003, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://mistral.ai/pricing#api-pricing" + }, + "mistral/mistral-ocr-2505-completion": { + "litellm_provider": "mistral", + "ocr_cost_per_page": 0.001, + "annotation_cost_per_page": 0.003, + "mode": "ocr", + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://mistral.ai/pricing#api-pricing" + }, + "mistral/magistral-medium-latest": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://mistral.ai/news/magistral", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/magistral-small-2506": { + "input_cost_per_token": 5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://mistral.ai/pricing#api-pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/magistral-small-latest": { + "input_cost_per_token": 5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://mistral.ai/pricing#api-pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/magistral-small-1-2-2509": { + "input_cost_per_token": 5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 40000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://mistral.ai/pricing#api-pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-embed": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding" + }, + "mistral/codestral-embed": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding" + }, + "mistral/codestral-embed-2505": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding" + }, + "mistral/mistral-large-2402": { + "input_cost_per_token": 4e-06, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-large-2407": { + "input_cost_per_token": 3e-06, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 9e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-large-2411": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-large-latest": { + "input_cost_per_token": 5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://docs.mistral.ai/models/mistral-large-3-25-12", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-large-3": { + "input_cost_per_token": 5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://docs.mistral.ai/models/mistral-large-3-25-12", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-large-2512": { + "input_cost_per_token": 5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://docs.mistral.ai/models/mistral-large-3-25-12", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-medium": { + "input_cost_per_token": 2.7e-06, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 8.1e-06, + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-medium-2312": { + "input_cost_per_token": 2.7e-06, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 8.1e-06, + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-medium-2505": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 131072, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-medium-latest": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-medium-3-1-2508": { + "input_cost_per_token": 4e-07, + "litellm_provider": "mistral", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://mistral.ai/news/mistral-medium-3", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-small": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/mistral-small-latest": { + "input_cost_per_token": 6e-08, + "litellm_provider": "mistral", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.8e-07, + "source": "https://mistral.ai/pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-small-3-2-2506": { + "input_cost_per_token": 6e-08, + "litellm_provider": "mistral", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.8e-07, + "source": "https://mistral.ai/pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/ministral-3-3b-2512": { + "input_cost_per_token": 1e-07, + "litellm_provider": "mistral", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://mistral.ai/pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/ministral-3-8b-2512": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://mistral.ai/pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/ministral-3-14b-2512": { + "input_cost_per_token": 2e-07, + "litellm_provider": "mistral", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://mistral.ai/pricing", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/mistral-tiny": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/open-codestral-mamba": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "source": "https://mistral.ai/technology/", + "supports_assistant_prefill": true, + "supports_tool_choice": true + }, + "mistral/open-mistral-7b": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/open-mistral-nemo": { + "input_cost_per_token": 3e-07, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://mistral.ai/technology/", + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/open-mistral-nemo-2407": { + "input_cost_per_token": 3e-07, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://mistral.ai/technology/", + "supports_assistant_prefill": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/open-mixtral-8x22b": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 65336, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/open-mixtral-8x7b": { + "input_cost_per_token": 7e-07, + "litellm_provider": "mistral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 7e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "mistral/pixtral-12b-2409": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/pixtral-large-2411": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "mistral/pixtral-large-latest": { + "input_cost_per_token": 2e-06, + "litellm_provider": "mistral", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot.kimi-k2-thinking": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "moonshotai.kimi-k2.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "moonshot/kimi-k2-0711-preview": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "moonshot/kimi-k2-0905-preview": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "moonshot/kimi-k2-turbo-preview": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 1.15e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 8e-06, + "source": "https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "moonshot/kimi-k2.5": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://platform.moonshot.ai/docs/guide/kimi-k2-5-quickstart", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true + }, + "moonshot/kimi-k2.6": { + "cache_read_input_token_cost": 1.6e-07, + "input_cost_per_token": 9.5e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://platform.kimi.ai/docs/pricing/chat-k26", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true + }, + "moonshot/kimi-latest": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/kimi-latest-128k": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/kimi-latest-32k": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/kimi-latest-8k": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 2e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/kimi-thinking-preview": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2", + "supports_vision": true + }, + "moonshot/kimi-k2-thinking": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "moonshot/kimi-k2-thinking-turbo": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 1.15e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 8e-06, + "source": "https://platform.moonshot.ai/docs/pricing/chat#generation-model-kimi-k2", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "moonshot/moonshot-v1-128k": { + "input_cost_per_token": 2e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "moonshot/moonshot-v1-128k-0430": { + "input_cost_per_token": 2e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "moonshot/moonshot-v1-128k-vision-preview": { + "input_cost_per_token": 2e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/moonshot-v1-32k": { + "input_cost_per_token": 1e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "moonshot/moonshot-v1-32k-0430": { + "input_cost_per_token": 1e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "moonshot/moonshot-v1-32k-vision-preview": { + "input_cost_per_token": 1e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/moonshot-v1-8k": { + "input_cost_per_token": 2e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "moonshot/moonshot-v1-8k-0430": { + "input_cost_per_token": 2e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "moonshot/moonshot-v1-8k-vision-preview": { + "input_cost_per_token": 2e-07, + "litellm_provider": "moonshot", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "moonshot/moonshot-v1-auto": { + "input_cost_per_token": 2e-06, + "litellm_provider": "moonshot", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://platform.moonshot.ai/docs/pricing", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "morph/morph-v3-fast": { + "input_cost_per_token": 8e-07, + "litellm_provider": "morph", + "max_input_tokens": 16000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_system_messages": true, + "supports_tool_choice": false, + "supports_vision": false + }, + "morph/morph-v3-large": { + "input_cost_per_token": 9e-07, + "litellm_provider": "morph", + "max_input_tokens": 16000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "mode": "chat", + "output_cost_per_token": 1.9e-06, + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_system_messages": true, + "supports_tool_choice": false, + "supports_vision": false + }, + "multimodalembedding": { + "input_cost_per_character": 2e-07, + "input_cost_per_image": 0.0001, + "input_cost_per_token": 8e-07, + "input_cost_per_video_per_second": 0.0005, + "input_cost_per_video_per_second_above_15s_interval": 0.002, + "input_cost_per_video_per_second_above_8s_interval": 0.001, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 768, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models", + "supported_endpoints": [ + "/v1/embeddings" + ], + "supported_modalities": [ + "text", + "image", + "video" + ] + }, + "multimodalembedding@001": { + "input_cost_per_character": 2e-07, + "input_cost_per_image": 0.0001, + "input_cost_per_token": 8e-07, + "input_cost_per_video_per_second": 0.0005, + "input_cost_per_video_per_second_above_15s_interval": 0.002, + "input_cost_per_video_per_second_above_8s_interval": 0.001, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 768, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models", + "supported_endpoints": [ + "/v1/embeddings" + ], + "supported_modalities": [ + "text", + "image", + "video" + ] + }, + "nscale/Qwen/QwQ-32B": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "nscale", + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/Qwen/Qwen2.5-Coder-32B-Instruct": { + "input_cost_per_token": 6e-08, + "litellm_provider": "nscale", + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/Qwen/Qwen2.5-Coder-3B-Instruct": { + "input_cost_per_token": 1e-08, + "litellm_provider": "nscale", + "mode": "chat", + "output_cost_per_token": 3e-08, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/Qwen/Qwen2.5-Coder-7B-Instruct": { + "input_cost_per_token": 1e-08, + "litellm_provider": "nscale", + "mode": "chat", + "output_cost_per_token": 3e-08, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/black-forest-labs/FLUX.1-schnell": { + "input_cost_per_pixel": 1.3e-09, + "litellm_provider": "nscale", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#image-models", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "nscale/deepseek-ai/DeepSeek-R1-Distill-Llama-70B": { + "input_cost_per_token": 3.75e-07, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.75/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 3.75e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/deepseek-ai/DeepSeek-R1-Distill-Llama-8B": { + "input_cost_per_token": 2.5e-08, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.05/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 2.5e-08, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B": { + "input_cost_per_token": 9e-08, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.18/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 9e-08, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B": { + "input_cost_per_token": 7e-08, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.14/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 7e-08, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.30/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B": { + "input_cost_per_token": 2e-07, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.40/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/meta-llama/Llama-3.1-8B-Instruct": { + "input_cost_per_token": 3e-08, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.06/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 3e-08, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/meta-llama/Llama-3.3-70B-Instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $0.40/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/meta-llama/Llama-4-Scout-17B-16E-Instruct": { + "input_cost_per_token": 9e-08, + "litellm_provider": "nscale", + "mode": "chat", + "output_cost_per_token": 2.9e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/mistralai/mixtral-8x22b-instruct-v0.1": { + "input_cost_per_token": 6e-07, + "litellm_provider": "nscale", + "metadata": { + "notes": "Pricing listed as $1.20/1M tokens total. Assumed 50/50 split for input/output." + }, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#chat-models" + }, + "nscale/stabilityai/stable-diffusion-xl-base-1.0": { + "input_cost_per_pixel": 3e-09, + "litellm_provider": "nscale", + "mode": "image_generation", + "output_cost_per_pixel": 0.0, + "source": "https://docs.nscale.com/docs/inference/serverless-models/current#image-models", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "nebius/deepseek-ai/DeepSeek-R1": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 8e-07, + "output_cost_per_token": 2.4e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/deepseek-ai/DeepSeek-R1-0528": { + "max_tokens": 164000, + "max_input_tokens": 164000, + "max_output_tokens": 164000, + "input_cost_per_token": 8e-07, + "output_cost_per_token": 2.4e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/deepseek-ai/DeepSeek-R1-Distill-Llama-70B": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 7.5e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/deepseek-ai/DeepSeek-V3": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 1.5e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/deepseek-ai/DeepSeek-V3-0324": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 1.5e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/google/gemma-3-27b-it": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 2e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "supports_vision": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/meta-llama/Llama-3.3-70B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/meta-llama/Llama-Guard-3-8B": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 6e-08, + "litellm_provider": "nebius", + "mode": "chat", + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/meta-llama/Meta-Llama-3.1-8B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 6e-08, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/meta-llama/Meta-Llama-3.1-70B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/meta-llama/Meta-Llama-3.1-405B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/mistralai/Mistral-Nemo-Instruct-2407": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1.2e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/NousResearch/Hermes-3-Llama-3.1-405B": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/nvidia/Llama-3.1-Nemotron-Ultra-253B-v1": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 1.8e-06, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/nvidia/Llama-3.3-Nemotron-Super-49B-v1": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen3-235B-A22B": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen3-32B": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 3e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen3-30B-A3B": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 3e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen3-14B": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 2.4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen3-4B": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 2.4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/QwQ-32B": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 4.5e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen2.5-72B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen2.5-32B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 2e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen2.5-Coder-7B": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-08, + "output_cost_per_token": 3e-08, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen2.5-VL-72B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "supports_vision": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen2-VL-72B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "nebius", + "mode": "chat", + "supports_function_calling": true, + "supports_vision": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/Qwen/Qwen2-VL-7B-Instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 6e-08, + "litellm_provider": "nebius", + "mode": "chat", + "supports_vision": true, + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/BAAI/bge-en-icl": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "input_cost_per_token": 1e-08, + "output_cost_per_token": 0.0, + "litellm_provider": "nebius", + "mode": "embedding", + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/BAAI/bge-multilingual-gemma2": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "input_cost_per_token": 1e-08, + "output_cost_per_token": 0.0, + "litellm_provider": "nebius", + "mode": "embedding", + "source": "https://nebius.com/prices-ai-studio" + }, + "nebius/intfloat/e5-mistral-7b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "input_cost_per_token": 1e-08, + "output_cost_per_token": 0.0, + "litellm_provider": "nebius", + "mode": "embedding", + "source": "https://nebius.com/prices-ai-studio" + }, + "nvidia.nemotron-nano-12b-v2": { + "input_cost_per_token": 2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_system_messages": true, + "supports_vision": true + }, + "nvidia.nemotron-nano-9b-v2": { + "input_cost_per_token": 6e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.3e-07, + "supports_system_messages": true + }, + "nvidia.nemotron-nano-3-30b": { + "input_cost_per_token": 6e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.4e-07, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_native_structured_output": true + }, + "nvidia.nemotron-super-3-120b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 256000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6.5e-07, + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "o1": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "o1-2024-12-17": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "o1-pro": { + "input_cost_per_token": 0.00015, + "input_cost_per_token_batches": 7.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 0.0006, + "output_cost_per_token_batches": 0.0003, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": false, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "o1-pro-2025-03-19": { + "input_cost_per_token": 0.00015, + "input_cost_per_token_batches": 7.5e-05, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 0.0006, + "output_cost_per_token_batches": 0.0003, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": false, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "o3": { + "cache_read_input_token_cost": 5e-07, + "cache_read_input_token_cost_flex": 2.5e-07, + "cache_read_input_token_cost_priority": 8.75e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_flex": 1e-06, + "input_cost_per_token_priority": 3.5e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "output_cost_per_token_flex": 4e-06, + "output_cost_per_token_priority": 1.4e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "o3-2025-04-16": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supported_endpoints": [ + "/v1/responses", + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "o3-deep-research": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_token": 1e-05, + "input_cost_per_token_batches": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 4e-05, + "output_cost_per_token_batches": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "o3-deep-research-2025-06-26": { + "cache_read_input_token_cost": 2.5e-06, + "input_cost_per_token": 1e-05, + "input_cost_per_token_batches": 5e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 4e-05, + "output_cost_per_token_batches": 2e-05, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "o3-mini": { + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "o3-mini-2025-01-31": { + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "o3-pro": { + "input_cost_per_token": 2e-05, + "input_cost_per_token_batches": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 8e-05, + "output_cost_per_token_batches": 4e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "o3-pro-2025-06-10": { + "input_cost_per_token": 2e-05, + "input_cost_per_token_batches": 1e-05, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 8e-05, + "output_cost_per_token_batches": 4e-05, + "supported_endpoints": [ + "/v1/responses", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "o4-mini": { + "cache_read_input_token_cost": 2.75e-07, + "cache_read_input_token_cost_flex": 1.375e-07, + "cache_read_input_token_cost_priority": 5e-07, + "input_cost_per_token": 1.1e-06, + "input_cost_per_token_flex": 5.5e-07, + "input_cost_per_token_priority": 2e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "output_cost_per_token_flex": 2.2e-06, + "output_cost_per_token_priority": 8e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "o4-mini-2025-04-16": { + "cache_read_input_token_cost": 2.75e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_service_tier": true, + "supports_vision": true, + "supports_web_search": true + }, + "o4-mini-deep-research": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 8e-06, + "output_cost_per_token_batches": 4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "o4-mini-deep-research-2025-06-26": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "openai", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "responses", + "output_cost_per_token": 8e-06, + "output_cost_per_token_batches": 4e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/batch", + "/v1/responses" + ], + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_native_streaming": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "oci/meta.llama-3.1-405b-instruct": { + "input_cost_per_token": 1.068e-05, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.068e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/meta.llama-3.2-90b-vision-instruct": { + "input_cost_per_token": 2e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false, + "supports_vision": true + }, + "oci/meta.llama-3.3-70b-instruct": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/meta.llama-4-maverick-17b-128e-instruct-fp8": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "oci", + "max_input_tokens": 512000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/meta.llama-4-scout-17b-16e-instruct": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "oci", + "max_input_tokens": 192000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-3": { + "input_cost_per_token": 3e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-3-fast": { + "input_cost_per_token": 5e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-3-mini": { + "input_cost_per_token": 3e-07, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-3-mini-fast": { + "input_cost_per_token": 6e-07, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-4": { + "input_cost_per_token": 3e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/cohere.command-latest": { + "input_cost_per_token": 1.56e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.56e-06, + "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/cohere.command-a-03-2025": { + "input_cost_per_token": 1.56e-06, + "litellm_provider": "oci", + "max_input_tokens": 256000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.56e-06, + "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/cohere.command-plus-latest": { + "input_cost_per_token": 1.56e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.56e-06, + "source": "https://www.oracle.com/cloud/ai/generative-ai/pricing/", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/cohere.command-a-reasoning-08-2025": { + "input_cost_per_token": 1.56e-06, + "litellm_provider": "oci", + "max_input_tokens": 256000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.56e-06, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/cohere.command-a-vision-07-2025": { + "input_cost_per_token": 1.56e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.56e-06, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false, + "supports_vision": true + }, + "oci/cohere.command-a-translate-08-2025": { + "input_cost_per_token": 9e-08, + "litellm_provider": "oci", + "max_input_tokens": 256000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 9e-08, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": false, + "supports_response_schema": false + }, + "oci/cohere.command-r-08-2024": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/cohere.command-r-plus-08-2024": { + "input_cost_per_token": 1.56e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.56e-06, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/meta.llama-3.2-11b-vision-instruct": { + "input_cost_per_token": 2e-06, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false, + "supports_vision": true + }, + "oci/meta.llama-3.1-70b-instruct": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/meta.llama-3.3-70b-instruct-fp8-dynamic": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-4-fast": { + "input_cost_per_token": 5e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-4.1-fast": { + "input_cost_per_token": 5e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-4.20": { + "input_cost_per_token": 3e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-4.20-multi-agent": { + "input_cost_per_token": 3e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/xai.grok-code-fast-1": { + "input_cost_per_token": 5e-06, + "litellm_provider": "oci", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": false + }, + "oci/google.gemini-2.5-pro": { + "input_cost_per_token": 1.25e-06, + "litellm_provider": "oci", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1e-05, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "oci/google.gemini-2.5-flash": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "oci", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "oci/google.gemini-2.5-flash-lite": { + "input_cost_per_token": 7.5e-08, + "litellm_provider": "oci", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "oci/cohere.embed-english-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing" + }, + "oci/cohere.embed-english-light-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 384, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing" + }, + "oci/cohere.embed-multilingual-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing" + }, + "oci/cohere.embed-multilingual-light-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 384, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing" + }, + "oci/cohere.embed-english-image-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_embedding_image_input": true + }, + "oci/cohere.embed-english-light-image-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 384, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_embedding_image_input": true + }, + "oci/cohere.embed-multilingual-light-image-v3.0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "oci", + "max_input_tokens": 512, + "max_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 384, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_embedding_image_input": true + }, + "oci/cohere.embed-v4.0": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "oci", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536, + "source": "https://www.oracle.com/artificial-intelligence/generative-ai/generative-ai-service/pricing", + "supports_embedding_image_input": true + }, + "ollama/codegeex4": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": false + }, + "ollama/codegemma": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "ollama/codellama": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "ollama/deepseek-coder-v2-base": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/deepseek-coder-v2-instruct": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/deepseek-coder-v2-lite-base": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/deepseek-coder-v2-lite-instruct": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/deepseek-v3.1:671b-cloud": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/gpt-oss:120b-cloud": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/gpt-oss:20b-cloud": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/internlm2_5-20b-chat": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/llama2": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/llama2-uncensored": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "ollama/llama2:13b": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/llama2:70b": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/llama2:7b": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/llama3": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/llama3.1": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/llama3:70b": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/llama3:8b": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "ollama/mistral": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "completion", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/mistral-7B-Instruct-v0.1": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/mistral-7B-Instruct-v0.2": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/mistral-large-instruct-2407": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/mixtral-8x22B-Instruct-v0.1": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/mixtral-8x7B-Instruct-v0.1": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/orca-mini": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "ollama/qwen3-coder:480b-cloud": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 0.0, + "supports_function_calling": true + }, + "ollama/vicuna": { + "input_cost_per_token": 0.0, + "litellm_provider": "ollama", + "max_input_tokens": 2048, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "omni-moderation-2024-09-26": { + "input_cost_per_token": 0.0, + "litellm_provider": "openai", + "max_input_tokens": 32768, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "moderation", + "output_cost_per_token": 0.0 + }, + "omni-moderation-latest": { + "input_cost_per_token": 0.0, + "litellm_provider": "openai", + "max_input_tokens": 32768, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "moderation", + "output_cost_per_token": 0.0 + }, + "openai.gpt-oss-120b-1:0": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "openai.gpt-oss-20b-1:0": { + "input_cost_per_token": 7e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "openai.gpt-oss-safeguard-120b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_system_messages": true + }, + "openai.gpt-oss-safeguard-20b": { + "input_cost_per_token": 7e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_system_messages": true + }, + "openrouter/anthropic/claude-3-haiku": { + "input_cost_per_image": 0.0004, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "openrouter", + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "max_input_tokens": 200000, + "max_output_tokens": 4096 + }, + "openrouter/anthropic/claude-3.5-sonnet": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-3.7-sonnet": { + "input_cost_per_image": 0.0048, + "input_cost_per_token": 3e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-opus-4": { + "input_cost_per_image": 0.0048, + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-opus-4.1": { + "input_cost_per_image": 0.0048, + "cache_creation_input_token_cost": 1.875e-05, + "cache_creation_input_token_cost_above_1hr": 3e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-sonnet-4": { + "input_cost_per_image": 0.0048, + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost": 3e-07, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-sonnet-4.6": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost": 3e-07, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "source": "https://openrouter.ai/anthropic/claude-sonnet-4.6", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_minimal_reasoning_effort": true + }, + "openrouter/anthropic/claude-opus-4.5": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-opus-4.6": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_minimal_reasoning_effort": true + }, + "openrouter/anthropic/claude-sonnet-4.5": { + "input_cost_per_image": 0.0048, + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 1000000, + "max_tokens": 1000000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "openrouter/anthropic/claude-haiku-4.5": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "max_tokens": 200000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346 + }, + "openrouter/anthropic/claude-opus-4.7": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346 + }, + "openrouter/bytedance/ui-tars-1.5-7b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 131072, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://openrouter.ai/api/v1/models/bytedance/ui-tars-1.5-7b", + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-chat": { + "input_cost_per_token": 1.4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.8e-07, + "supports_prompt_caching": true, + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-chat-v3-0324": { + "input_cost_per_token": 1.4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.8e-07, + "supports_prompt_caching": true, + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-chat-v3.1": { + "input_cost_per_token": 2e-07, + "input_cost_per_token_cache_hit": 2e-08, + "litellm_provider": "openrouter", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 8e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-v3.2": { + "input_cost_per_token": 2.8e-07, + "input_cost_per_token_cache_hit": 2.8e-08, + "litellm_provider": "openrouter", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-v3.2-exp": { + "input_cost_per_token": 2e-07, + "input_cost_per_token_cache_hit": 2e-08, + "litellm_provider": "openrouter", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": false, + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-r1": { + "input_cost_per_token": 5.5e-07, + "input_cost_per_token_cache_hit": 1.4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 65336, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.19e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/deepseek/deepseek-r1-0528": { + "input_cost_per_token": 5e-07, + "input_cost_per_token_cache_hit": 1.4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 65336, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.15e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/google/gemini-2.0-flash-001": { + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/google/gemini-2.5-flash": { + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 3e-07, + "litellm_provider": "openrouter", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/google/gemini-2.5-pro": { + "input_cost_per_audio_token": 7e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openrouter", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 30, + "max_tokens": 8192, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/google/gemini-3-pro-preview": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "openrouter", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "output_cost_per_token_batches": 6e-06, + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true + }, + "openrouter/google/gemini-3-flash-preview": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 5e-07, + "litellm_provider": "openrouter", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 3e-06, + "output_cost_per_token": 3e-06, + "rpm": 2000, + "source": "https://ai.google.dev/pricing/gemini-3", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 800000 + }, + "openrouter/google/gemini-3.1-flash-lite-preview": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_per_audio_token": 5e-08, + "input_cost_per_audio_token": 5e-07, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "openrouter", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 1.5e-06, + "output_cost_per_token": 1.5e-06, + "rpm": 2000, + "source": "https://ai.google.dev/pricing/gemini-3", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_audio_output": false, + "supports_code_execution": true, + "supports_file_search": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 800000 + }, + "openrouter/google/gemini-3.1-pro-preview": { + "cache_read_input_token_cost": 2e-07, + "cache_read_input_token_cost_above_200k_tokens": 4e-07, + "cache_creation_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 2e-06, + "input_cost_per_token_above_200k_tokens": 4e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_above_200k_tokens": 1.8e-05, + "source": "https://openrouter.ai/google/gemini-3.1-pro-preview", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/gryphe/mythomax-l2-13b": { + "input_cost_per_token": 1.875e-06, + "litellm_provider": "openrouter", + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.875e-06, + "supports_tool_choice": true + }, + "openrouter/mancer/weaver": { + "input_cost_per_token": 5.625e-06, + "litellm_provider": "openrouter", + "max_tokens": 2000, + "mode": "chat", + "output_cost_per_token": 5.625e-06, + "supports_tool_choice": true, + "max_input_tokens": 8000, + "max_output_tokens": 2000 + }, + "openrouter/meta-llama/llama-3-70b-instruct": { + "input_cost_per_token": 5.9e-07, + "litellm_provider": "openrouter", + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 7.9e-07, + "supports_tool_choice": true, + "max_input_tokens": 8192, + "max_output_tokens": 8000 + }, + "openrouter/minimax/minimax-m2": { + "input_cost_per_token": 2.55e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 204800, + "max_output_tokens": 204800, + "max_tokens": 204800, + "mode": "chat", + "output_cost_per_token": 1.02e-06, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/mistralai/devstral-2512": { + "input_cost_per_image": 0, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_prompt_caching": false, + "supports_tool_choice": true, + "supports_vision": false + }, + "openrouter/mistralai/ministral-3b-2512": { + "input_cost_per_image": 0, + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_prompt_caching": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/mistralai/ministral-8b-2512": { + "input_cost_per_image": 0, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_function_calling": true, + "supports_prompt_caching": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/mistralai/ministral-14b-2512": { + "input_cost_per_image": 0, + "input_cost_per_token": 2e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_function_calling": true, + "supports_prompt_caching": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/mistralai/mistral-large-2512": { + "input_cost_per_image": 0, + "input_cost_per_token": 5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_prompt_caching": false, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/mistralai/mistral-7b-instruct": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "openrouter", + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 1.3e-07, + "supports_tool_choice": true, + "max_input_tokens": 32768, + "max_output_tokens": 8191 + }, + "openrouter/mistralai/mistral-large": { + "input_cost_per_token": 8e-06, + "litellm_provider": "openrouter", + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2.4e-05, + "supports_tool_choice": true, + "max_input_tokens": 128000, + "max_output_tokens": 8191 + }, + "openrouter/mistralai/mistral-small-3.1-24b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_tool_choice": true, + "max_input_tokens": 131072, + "max_output_tokens": 131072 + }, + "openrouter/mistralai/mistral-small-3.2-24b-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_tool_choice": true, + "max_input_tokens": 128000, + "max_output_tokens": 128000 + }, + "openrouter/mistralai/mixtral-8x22b-instruct": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "openrouter", + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 6.5e-07, + "supports_tool_choice": true, + "max_input_tokens": 65536, + "max_output_tokens": 65536 + }, + "openrouter/moonshotai/kimi-k2.5": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://openrouter.ai/moonshotai/kimi-k2.5", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true + }, + "openrouter/openai/gpt-3.5-turbo": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "openrouter", + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_tool_choice": true, + "max_input_tokens": 16385, + "max_output_tokens": 4096 + }, + "openrouter/openai/gpt-3.5-turbo-16k": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openrouter", + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_tool_choice": true, + "max_input_tokens": 16385, + "max_output_tokens": 4096 + }, + "openrouter/openai/gpt-4": { + "input_cost_per_token": 3e-05, + "litellm_provider": "openrouter", + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_tool_choice": true, + "max_input_tokens": 8191, + "max_output_tokens": 4096 + }, + "openrouter/openai/gpt-4.1": { + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-4.1-mini": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-4.1-nano": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-4o": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-4o-2024-05-13": { + "input_cost_per_token": 5e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-5-chat": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-5-codex": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-5.2-codex": { + "cache_read_input_token_cost": 1.75e-07, + "input_cost_per_token": 1.75e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-5": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-5-mini": { + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-5-nano": { + "cache_read_input_token_cost": 5e-09, + "input_cost_per_token": 5e-08, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-5.1-codex-max": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 400000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "source": "https://openrouter.ai/openai/gpt-5.1-codex-max", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text" + ], + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-5.2": { + "input_cost_per_image": 0, + "cache_read_input_token_cost": 1.75e-07, + "input_cost_per_token": 1.75e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-5.2-chat": { + "input_cost_per_image": 0, + "cache_read_input_token_cost": 1.75e-07, + "input_cost_per_token": 1.75e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-05, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-5.2-pro": { + "input_cost_per_image": 0, + "input_cost_per_token": 2.1e-05, + "litellm_provider": "openrouter", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 0.000168, + "supports_function_calling": true, + "supports_prompt_caching": false, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/gpt-oss-120b": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-07, + "source": "https://openrouter.ai/openai/gpt-oss-120b", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "openrouter/openai/gpt-oss-20b": { + "input_cost_per_token": 2e-08, + "litellm_provider": "openrouter", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://openrouter.ai/openai/gpt-oss-20b", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "openrouter/openai/o1": { + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/openai/o3-mini": { + "input_cost_per_token": 1.1e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "openrouter/openai/o3-mini-high": { + "input_cost_per_token": 1.1e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "openrouter/qwen/qwen-2.5-coder-32b-instruct": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 33792, + "max_output_tokens": 33792, + "max_tokens": 33792, + "mode": "chat", + "output_cost_per_token": 1.8e-07, + "supports_tool_choice": true + }, + "openrouter/qwen/qwen-vl-plus": { + "input_cost_per_token": 2.1e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 8192, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 6.3e-07, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/qwen/qwen3-coder": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262100, + "max_output_tokens": 262100, + "max_tokens": 262100, + "mode": "chat", + "output_cost_per_token": 9.5e-07, + "source": "https://openrouter.ai/qwen/qwen3-coder", + "supports_tool_choice": true, + "supports_function_calling": true + }, + "openrouter/qwen/qwen3-coder-plus": { + "input_cost_per_token": 1e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 997952, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://openrouter.ai/qwen/qwen3-coder-plus", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/qwen/qwen3-235b-a22b-2507": { + "input_cost_per_token": 7.1e-08, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://openrouter.ai/qwen/qwen3-235b-a22b-2507", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "openrouter/qwen/qwen3-235b-a22b-thinking-2507": { + "input_cost_per_token": 1.1e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://openrouter.ai/qwen/qwen3-235b-a22b-thinking-2507", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/qwen/qwen3.5-35b-a3b": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://openrouter.ai/qwen/qwen3.5-35b-a3b", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/qwen/qwen3.5-27b": { + "input_cost_per_token": 3e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 2.4e-06, + "source": "https://openrouter.ai/qwen/qwen3.5-27b", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/qwen/qwen3.5-122b-a10b": { + "input_cost_per_token": 4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://openrouter.ai/qwen/qwen3.5-122b-a10b", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/qwen/qwen3.5-flash-02-23": { + "input_cost_per_token": 1e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 4e-07, + "source": "https://openrouter.ai/qwen/qwen3.5-flash-02-23", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/qwen/qwen3.5-plus-02-15": { + "input_cost_per_token": 4e-07, + "input_cost_per_token_above_256k_tokens": 5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 1000000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 2.4e-06, + "output_cost_per_token_above_256k_tokens": 3e-06, + "source": "https://openrouter.ai/qwen/qwen3.5-plus-02-15", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/qwen/qwen3.5-397b-a17b": { + "input_cost_per_token": 6e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "source": "https://openrouter.ai/qwen/qwen3.5-397b-a17b", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "openrouter/switchpoint/router": { + "input_cost_per_token": 8.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 3.4e-06, + "source": "https://openrouter.ai/switchpoint/router", + "supports_tool_choice": true + }, + "openrouter/undi95/remm-slerp-l2-13b": { + "input_cost_per_token": 1.875e-06, + "litellm_provider": "openrouter", + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.875e-06, + "supports_tool_choice": true, + "max_input_tokens": 6144, + "max_output_tokens": 4096 + }, + "openrouter/x-ai/grok-4": { + "input_cost_per_token": 3e-06, + "litellm_provider": "openrouter", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://openrouter.ai/x-ai/grok-4", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "openrouter/z-ai/glm-4.6": { + "input_cost_per_token": 4e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 202800, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 1.75e-06, + "source": "https://openrouter.ai/z-ai/glm-4.6", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/z-ai/glm-4.6:exacto": { + "input_cost_per_token": 4.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 202800, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 1.9e-06, + "source": "https://openrouter.ai/z-ai/glm-4.6:exacto", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/xiaomi/mimo-v2-flash": { + "input_cost_per_token": 9e-08, + "output_cost_per_token": 2.9e-07, + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 0.0, + "litellm_provider": "openrouter", + "max_input_tokens": 262144, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_vision": false, + "supports_prompt_caching": false + }, + "openrouter/z-ai/glm-4.7": { + "input_cost_per_token": 4e-07, + "output_cost_per_token": 1.5e-06, + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 0.0, + "litellm_provider": "openrouter", + "max_input_tokens": 202752, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_vision": true, + "supports_prompt_caching": false, + "supports_assistant_prefill": true + }, + "openrouter/z-ai/glm-4.7-flash": { + "input_cost_per_token": 7e-08, + "output_cost_per_token": 4e-07, + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 0.0, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_vision": true, + "supports_prompt_caching": false + }, + "openrouter/z-ai/glm-5": { + "input_cost_per_token": 8e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 202752, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.56e-06, + "source": "https://openrouter.ai/z-ai/glm-5", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "openrouter/minimax/minimax-m2.1": { + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1.2e-06, + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 0.0, + "litellm_provider": "openrouter", + "max_input_tokens": 204000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_vision": true, + "supports_prompt_caching": false, + "supports_computer_use": false + }, + "openrouter/minimax/minimax-m2.5": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.1e-06, + "cache_read_input_token_cost": 1.5e-07, + "litellm_provider": "openrouter", + "max_input_tokens": 196608, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "source": "https://openrouter.ai/minimax/minimax-m2.5", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_vision": false, + "supports_prompt_caching": true, + "supports_computer_use": false + }, + "openrouter/openrouter/auto": { + "input_cost_per_token": 0, + "output_cost_per_token": 0, + "litellm_provider": "openrouter", + "max_input_tokens": 2000000, + "max_tokens": 2000000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_vision": true, + "supports_audio_input": true, + "supports_video_input": true + }, + "openrouter/openrouter/free": { + "input_cost_per_token": 0, + "output_cost_per_token": 0, + "litellm_provider": "openrouter", + "max_input_tokens": 200000, + "max_tokens": 200000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_vision": true + }, + "openrouter/openrouter/bodybuilder": { + "input_cost_per_token": 0, + "output_cost_per_token": 0, + "litellm_provider": "openrouter", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "chat" + }, + "ovhcloud/DeepSeek-R1-Distill-Llama-70B": { + "input_cost_per_token": 6.7e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 6.7e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/deepseek-r1-distill-llama-70b", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "ovhcloud/Llama-3.1-8B-Instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/llama-3-1-8b-instruct", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "ovhcloud/Meta-Llama-3_1-70B-Instruct": { + "input_cost_per_token": 6.7e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 6.7e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-1-70b-instruct", + "supports_function_calling": false, + "supports_response_schema": false, + "supports_tool_choice": false + }, + "ovhcloud/Meta-Llama-3_3-70B-Instruct": { + "input_cost_per_token": 6.7e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 6.7e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/meta-llama-3-3-70b-instruct", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "ovhcloud/Mistral-7B-Instruct-v0.3": { + "input_cost_per_token": 1e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 127000, + "max_output_tokens": 127000, + "max_tokens": 127000, + "mode": "chat", + "output_cost_per_token": 1e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-7b-instruct-v0-3", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "ovhcloud/Mistral-Nemo-Instruct-2407": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 118000, + "max_output_tokens": 118000, + "max_tokens": 118000, + "mode": "chat", + "output_cost_per_token": 1.3e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-nemo-instruct-2407", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "ovhcloud/Mistral-Small-3.2-24B-Instruct-2506": { + "input_cost_per_token": 9e-08, + "litellm_provider": "ovhcloud", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.8e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/mistral-small-3-2-24b-instruct-2506", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "ovhcloud/Mixtral-8x7B-Instruct-v0.1": { + "input_cost_per_token": 6.3e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 6.3e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/mixtral-8x7b-instruct-v0-1", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "ovhcloud/Qwen2.5-Coder-32B-Instruct": { + "input_cost_per_token": 8.7e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 8.7e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/qwen2-5-coder-32b-instruct", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "ovhcloud/Qwen2.5-VL-72B-Instruct": { + "input_cost_per_token": 9.1e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 9.1e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/qwen2-5-vl-72b-instruct", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "ovhcloud/Qwen3-32B": { + "input_cost_per_token": 8e-08, + "litellm_provider": "ovhcloud", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 2.3e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/qwen3-32b", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "ovhcloud/gpt-oss-120b": { + "input_cost_per_token": 8e-08, + "litellm_provider": "ovhcloud", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/gpt-oss-120b", + "supports_function_calling": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "ovhcloud/gpt-oss-20b": { + "input_cost_per_token": 4e-08, + "litellm_provider": "ovhcloud", + "max_input_tokens": 131000, + "max_output_tokens": 131000, + "max_tokens": 131000, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/gpt-oss-20b", + "supports_function_calling": false, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "ovhcloud/llava-v1.6-mistral-7b-hf": { + "input_cost_per_token": 2.9e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 2.9e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/llava-next-mistral-7b", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "ovhcloud/mamba-codestral-7B-v0.1": { + "input_cost_per_token": 1.9e-07, + "litellm_provider": "ovhcloud", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.9e-07, + "source": "https://endpoints.ai.cloud.ovh.net/models/mamba-codestral-7b-v0-1", + "supports_function_calling": false, + "supports_response_schema": true, + "supports_tool_choice": false + }, + "palm/chat-bison": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "palm", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "palm/chat-bison-001": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "palm", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "palm/text-bison": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "palm", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "completion", + "output_cost_per_token": 1.25e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "palm/text-bison-001": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "palm", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "completion", + "output_cost_per_token": 1.25e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "palm/text-bison-safety-off": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "palm", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "completion", + "output_cost_per_token": 1.25e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "palm/text-bison-safety-recitation-off": { + "input_cost_per_token": 1.25e-07, + "litellm_provider": "palm", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "completion", + "output_cost_per_token": 1.25e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "parallel_ai/search": { + "input_cost_per_query": 0.004, + "litellm_provider": "parallel_ai", + "mode": "search" + }, + "parallel_ai/search-pro": { + "input_cost_per_query": 0.009, + "litellm_provider": "parallel_ai", + "mode": "search" + }, + "perplexity/codellama-34b-instruct": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "perplexity", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.4e-06 + }, + "perplexity/codellama-70b-instruct": { + "input_cost_per_token": 7e-07, + "litellm_provider": "perplexity", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2.8e-06 + }, + "perplexity/llama-2-70b-chat": { + "input_cost_per_token": 7e-07, + "litellm_provider": "perplexity", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-06 + }, + "perplexity/llama-3.1-70b-instruct": { + "input_cost_per_token": 1e-06, + "litellm_provider": "perplexity", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-06 + }, + "perplexity/llama-3.1-8b-instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "perplexity", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2e-07 + }, + "perplexity/mistral-7b-instruct": { + "input_cost_per_token": 7e-08, + "litellm_provider": "perplexity", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-07 + }, + "perplexity/mixtral-8x7b-instruct": { + "input_cost_per_token": 7e-08, + "litellm_provider": "perplexity", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-07 + }, + "perplexity/pplx-70b-chat": { + "input_cost_per_token": 7e-07, + "litellm_provider": "perplexity", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-06 + }, + "perplexity/pplx-70b-online": { + "input_cost_per_request": 0.005, + "input_cost_per_token": 0.0, + "litellm_provider": "perplexity", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-06 + }, + "perplexity/pplx-7b-chat": { + "input_cost_per_token": 7e-08, + "litellm_provider": "perplexity", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.8e-07 + }, + "perplexity/pplx-7b-online": { + "input_cost_per_request": 0.005, + "input_cost_per_token": 0.0, + "litellm_provider": "perplexity", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.8e-07 + }, + "perplexity/sonar": { + "input_cost_per_token": 1e-06, + "litellm_provider": "perplexity", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-06, + "search_context_cost_per_query": { + "search_context_size_high": 0.012, + "search_context_size_low": 0.005, + "search_context_size_medium": 0.008 + }, + "supports_web_search": true + }, + "perplexity/sonar-deep-research": { + "citation_cost_per_token": 2e-06, + "input_cost_per_token": 2e-06, + "litellm_provider": "perplexity", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_reasoning_token": 3e-06, + "output_cost_per_token": 8e-06, + "search_context_cost_per_query": { + "search_context_size_high": 0.005, + "search_context_size_low": 0.005, + "search_context_size_medium": 0.005 + }, + "supports_reasoning": true, + "supports_web_search": true + }, + "perplexity/sonar-medium-chat": { + "input_cost_per_token": 6e-07, + "litellm_provider": "perplexity", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.8e-06 + }, + "perplexity/sonar-medium-online": { + "input_cost_per_request": 0.005, + "input_cost_per_token": 0, + "litellm_provider": "perplexity", + "max_input_tokens": 12000, + "max_output_tokens": 12000, + "max_tokens": 12000, + "mode": "chat", + "output_cost_per_token": 1.8e-06 + }, + "perplexity/sonar-pro": { + "input_cost_per_token": 3e-06, + "litellm_provider": "perplexity", + "max_input_tokens": 200000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.014, + "search_context_size_low": 0.006, + "search_context_size_medium": 0.01 + }, + "supports_web_search": true + }, + "perplexity/sonar-reasoning": { + "input_cost_per_token": 1e-06, + "litellm_provider": "perplexity", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "search_context_cost_per_query": { + "search_context_size_high": 0.014, + "search_context_size_low": 0.005, + "search_context_size_medium": 0.008 + }, + "supports_reasoning": true, + "supports_web_search": true + }, + "perplexity/sonar-reasoning-pro": { + "input_cost_per_token": 2e-06, + "litellm_provider": "perplexity", + "max_input_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "search_context_cost_per_query": { + "search_context_size_high": 0.014, + "search_context_size_low": 0.006, + "search_context_size_medium": 0.01 + }, + "supports_reasoning": true, + "supports_web_search": true + }, + "perplexity/sonar-small-chat": { + "input_cost_per_token": 7e-08, + "litellm_provider": "perplexity", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2.8e-07 + }, + "perplexity/sonar-small-online": { + "input_cost_per_request": 0.005, + "input_cost_per_token": 0, + "litellm_provider": "perplexity", + "max_input_tokens": 12000, + "max_output_tokens": 12000, + "max_tokens": 12000, + "mode": "chat", + "output_cost_per_token": 2.8e-07 + }, + "publicai/swiss-ai/apertus-8b-instruct": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": false, + "supports_tool_choice": false + }, + "publicai/swiss-ai/apertus-70b-instruct": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": false, + "supports_tool_choice": false + }, + "publicai/aisingapore/Gemma-SEA-LION-v4-27B-IT": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "publicai/BSC-LT/salamandra-7b-instruct-tools-16k": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "publicai/BSC-LT/ALIA-40b-instruct_Q8_0": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "publicai/allenai/Olmo-3-7B-Instruct": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "perplexity/preset/fast-search": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_preset": true, + "supports_function_calling": true + }, + "perplexity/preset/pro-search": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_preset": true, + "supports_function_calling": true + }, + "perplexity/preset/deep-research": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_preset": true, + "supports_function_calling": true + }, + "perplexity/preset/advanced-deep-research": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_preset": true, + "supports_function_calling": true + }, + "perplexity/openai/gpt-5.2": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": true, + "supports_function_calling": true + }, + "perplexity/openai/gpt-5.1": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/openai/gpt-5-mini": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/anthropic/claude-opus-4-6": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/anthropic/claude-opus-4-7": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/anthropic/claude-opus-4-5": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/anthropic/claude-sonnet-4-5": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/anthropic/claude-haiku-4-5": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/google/gemini-3-pro-preview": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/google/gemini-3-flash-preview": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/google/gemini-2.5-pro": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/google/gemini-2.5-flash": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/xai/grok-4-1-fast-non-reasoning": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/perplexity/sonar": { + "litellm_provider": "perplexity", + "mode": "responses", + "supports_web_search": true, + "supports_reasoning": false, + "supports_function_calling": true + }, + "perplexity/pplx-embed-v1-0.6b": { + "input_cost_per_token": 4e-09, + "litellm_provider": "perplexity", + "max_input_tokens": 32768, + "max_tokens": 32768, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1024, + "source": "https://docs.perplexity.ai/docs/embeddings/quickstart" + }, + "perplexity/pplx-embed-v1-4b": { + "input_cost_per_token": 3e-08, + "litellm_provider": "perplexity", + "max_input_tokens": 32768, + "max_tokens": 32768, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 2560, + "source": "https://docs.perplexity.ai/docs/embeddings/quickstart" + }, + "publicai/aisingapore/Qwen-SEA-LION-v4-32B-IT": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "publicai/allenai/Olmo-3-7B-Think": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true + }, + "publicai/allenai/Olmo-3-32B-Think": { + "input_cost_per_token": 0.0, + "litellm_provider": "publicai", + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://platform.publicai.co/docs", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true + }, + "qwen.qwen3-coder-480b-a35b-v1:0": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 262000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1.8e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_native_structured_output": true + }, + "qwen.qwen3-235b-a22b-2507-v1:0": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 262144, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 8.8e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_native_structured_output": true + }, + "qwen.qwen3-coder-30b-a3b-v1:0": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 262144, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_native_structured_output": true + }, + "qwen.qwen3-32b-v1:0": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_native_structured_output": true + }, + "qwen.qwen3-next-80b-a3b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_native_structured_output": true + }, + "qwen.qwen3-vl-235b-a22b": { + "input_cost_per_token": 5.3e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.66e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_vision": true, + "supports_native_structured_output": true + }, + "qwen.qwen3-coder-next": { + "input_cost_per_token": 5e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "recraft/recraftv2": { + "litellm_provider": "recraft", + "mode": "image_generation", + "output_cost_per_image": 0.022, + "source": "https://www.recraft.ai/docs#pricing", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "recraft/recraftv3": { + "litellm_provider": "recraft", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://www.recraft.ai/docs#pricing", + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "replicate/meta/llama-2-13b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5e-07, + "supports_tool_choice": true + }, + "replicate/meta/llama-2-13b-chat": { + "input_cost_per_token": 1e-07, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5e-07, + "supports_tool_choice": true + }, + "replicate/meta/llama-2-70b": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_tool_choice": true + }, + "replicate/meta/llama-2-70b-chat": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_tool_choice": true + }, + "replicate/meta/llama-2-7b": { + "input_cost_per_token": 5e-08, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_tool_choice": true + }, + "replicate/meta/llama-2-7b-chat": { + "input_cost_per_token": 5e-08, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_tool_choice": true + }, + "replicate/meta/llama-3-70b": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "replicate", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_tool_choice": true + }, + "replicate/meta/llama-3-70b-instruct": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "replicate", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.75e-06, + "supports_tool_choice": true + }, + "replicate/meta/llama-3-8b": { + "input_cost_per_token": 5e-08, + "litellm_provider": "replicate", + "max_input_tokens": 8086, + "max_output_tokens": 8086, + "max_tokens": 8086, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_tool_choice": true + }, + "replicate/meta/llama-3-8b-instruct": { + "input_cost_per_token": 5e-08, + "litellm_provider": "replicate", + "max_input_tokens": 8086, + "max_output_tokens": 8086, + "max_tokens": 8086, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_tool_choice": true + }, + "replicate/mistralai/mistral-7b-instruct-v0.2": { + "input_cost_per_token": 5e-08, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_tool_choice": true + }, + "replicate/mistralai/mistral-7b-v0.1": { + "input_cost_per_token": 5e-08, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2.5e-07, + "supports_tool_choice": true + }, + "replicate/mistralai/mixtral-8x7b-instruct-v0.1": { + "input_cost_per_token": 3e-07, + "litellm_provider": "replicate", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-06, + "supports_tool_choice": true + }, + "replicate/openai/gpt-5": { + "input_cost_per_token": 1.25e-06, + "output_cost_per_token": 1e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicateopenai/gpt-oss-20b": { + "input_cost_per_token": 9e-08, + "output_cost_per_token": 3.6e-07, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/anthropic/claude-4.5-haiku": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 5e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_prompt_caching": true + }, + "replicate/ibm-granite/granite-3.3-8b-instruct": { + "input_cost_per_token": 3e-08, + "output_cost_per_token": 2.5e-07, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/openai/gpt-4o": { + "input_cost_per_token": 2.5e-06, + "output_cost_per_token": 1e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_audio_input": true, + "supports_audio_output": true + }, + "replicate/openai/o4-mini": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 4e-06, + "output_cost_per_reasoning_token": 4e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_reasoning": true, + "supports_system_messages": true + }, + "replicate/openai/o1-mini": { + "input_cost_per_token": 1.1e-06, + "output_cost_per_token": 4.4e-06, + "output_cost_per_reasoning_token": 4.4e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_reasoning": true, + "supports_system_messages": true + }, + "replicate/openai/o1": { + "input_cost_per_token": 1.5e-05, + "output_cost_per_token": 6e-05, + "output_cost_per_reasoning_token": 6e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_reasoning": true, + "supports_system_messages": true + }, + "replicate/openai/gpt-4o-mini": { + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicate/qwen/qwen3-235b-a22b-instruct-2507": { + "input_cost_per_token": 2.64e-07, + "output_cost_per_token": 1.06e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/anthropic/claude-4-sonnet": { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_prompt_caching": true + }, + "replicate/deepseek-ai/deepseek-v3": { + "input_cost_per_token": 1.45e-06, + "output_cost_per_token": 1.45e-06, + "litellm_provider": "replicate", + "mode": "chat", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/anthropic/claude-3.7-sonnet": { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_prompt_caching": true + }, + "replicate/anthropic/claude-3.5-haiku": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 5e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_prompt_caching": true + }, + "replicate/anthropic/claude-3.5-sonnet": { + "input_cost_per_token": 3.75e-06, + "output_cost_per_token": 1.875e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_prompt_caching": true + }, + "replicate/google/gemini-3-pro": { + "input_cost_per_token": 2e-06, + "output_cost_per_token": 1.2e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicate/anthropic/claude-4.5-sonnet": { + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1.5e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true, + "supports_prompt_caching": true + }, + "replicate/openai/gpt-4.1": { + "input_cost_per_token": 2e-06, + "output_cost_per_token": 8e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicate/openai/gpt-4.1-nano": { + "input_cost_per_token": 1e-07, + "output_cost_per_token": 4e-07, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/openai/gpt-4.1-mini": { + "input_cost_per_token": 4e-07, + "output_cost_per_token": 1.6e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicate/openai/gpt-5-nano": { + "input_cost_per_token": 5e-08, + "output_cost_per_token": 4e-07, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/openai/gpt-5-mini": { + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 2e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicate/google/gemini-2.5-flash": { + "input_cost_per_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "replicate/openai/gpt-oss-120b": { + "input_cost_per_token": 1.8e-07, + "output_cost_per_token": 7.2e-07, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/deepseek-ai/deepseek-v3.1": { + "input_cost_per_token": 6.72e-07, + "output_cost_per_token": 2.016e-06, + "litellm_provider": "replicate", + "mode": "chat", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true + }, + "replicate/xai/grok-4": { + "input_cost_per_token": 7.2e-06, + "output_cost_per_token": 3.6e-05, + "litellm_provider": "replicate", + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true + }, + "replicate/deepseek-ai/deepseek-r1": { + "input_cost_per_token": 3.75e-06, + "output_cost_per_token": 1e-05, + "output_cost_per_reasoning_token": 1e-05, + "litellm_provider": "replicate", + "mode": "chat", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_reasoning": true, + "supports_system_messages": true + }, + "rerank-english-v2.0": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "rerank-english-v3.0": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "rerank-multilingual-v2.0": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "rerank-multilingual-v3.0": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "rerank-v3.5": { + "input_cost_per_query": 0.002, + "input_cost_per_token": 0.0, + "litellm_provider": "cohere", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_query_tokens": 2048, + "max_tokens": 4096, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "nvidia_nim/nvidia/nv-rerankqa-mistral-4b-v3": { + "input_cost_per_query": 0.0, + "input_cost_per_token": 0.0, + "litellm_provider": "nvidia_nim", + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "nvidia_nim/nvidia/llama-3_2-nv-rerankqa-1b-v2": { + "input_cost_per_query": 0.0, + "input_cost_per_token": 0.0, + "litellm_provider": "nvidia_nim", + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "nvidia_nim/ranking/nvidia/llama-3.2-nv-rerankqa-1b-v2": { + "input_cost_per_query": 0.0, + "input_cost_per_token": 0.0, + "litellm_provider": "nvidia_nim", + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "sagemaker/meta-textgeneration-llama-2-13b": { + "input_cost_per_token": 0.0, + "litellm_provider": "sagemaker", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "sagemaker/meta-textgeneration-llama-2-13b-f": { + "input_cost_per_token": 0.0, + "litellm_provider": "sagemaker", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "sagemaker/meta-textgeneration-llama-2-70b": { + "input_cost_per_token": 0.0, + "litellm_provider": "sagemaker", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "sagemaker/meta-textgeneration-llama-2-70b-b-f": { + "input_cost_per_token": 0.0, + "litellm_provider": "sagemaker", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "sagemaker/meta-textgeneration-llama-2-7b": { + "input_cost_per_token": 0.0, + "litellm_provider": "sagemaker", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "completion", + "output_cost_per_token": 0.0 + }, + "sagemaker/meta-textgeneration-llama-2-7b-f": { + "input_cost_per_token": 0.0, + "litellm_provider": "sagemaker", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "sambanova/DeepSeek-R1": { + "input_cost_per_token": 5e-06, + "litellm_provider": "sambanova", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 7e-06, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/DeepSeek-R1-Distill-Llama-70B": { + "input_cost_per_token": 7e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.4e-06, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/DeepSeek-V3-0324": { + "input_cost_per_token": 3e-06, + "litellm_provider": "sambanova", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4.5e-06, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "sambanova/Llama-4-Maverick-17B-128E-Instruct": { + "input_cost_per_token": 6.3e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "metadata": { + "notes": "For vision models, images are converted to 6432 input tokens and are billed at that amount" + }, + "mode": "chat", + "output_cost_per_token": 1.8e-06, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "sambanova/Llama-4-Scout-17B-16E-Instruct": { + "input_cost_per_token": 4e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "metadata": { + "notes": "For vision models, images are converted to 6432 input tokens and are billed at that amount" + }, + "mode": "chat", + "output_cost_per_token": 7e-07, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "sambanova/Meta-Llama-3.1-405B-Instruct": { + "input_cost_per_token": 5e-06, + "litellm_provider": "sambanova", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "sambanova/Meta-Llama-3.1-8B-Instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "sambanova/Meta-Llama-3.2-1B-Instruct": { + "input_cost_per_token": 4e-08, + "litellm_provider": "sambanova", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 8e-08, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/Meta-Llama-3.2-3B-Instruct": { + "input_cost_per_token": 8e-08, + "litellm_provider": "sambanova", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.6e-07, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/Meta-Llama-3.3-70B-Instruct": { + "input_cost_per_token": 6e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "sambanova/Meta-Llama-Guard-3-8B": { + "input_cost_per_token": 3e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/QwQ-32B": { + "input_cost_per_token": 5e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-06, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/Qwen2-Audio-7B-Instruct": { + "input_cost_per_token": 5e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 0.0001, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_audio_input": true + }, + "sambanova/Qwen3-32B": { + "input_cost_per_token": 4e-07, + "litellm_provider": "sambanova", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 8e-07, + "source": "https://cloud.sambanova.ai/plans/pricing", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "sambanova/DeepSeek-V3.1": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 3e-06, + "output_cost_per_token": 4.5e-06, + "litellm_provider": "sambanova", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "sambanova/gpt-oss-120b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 3e-06, + "output_cost_per_token": 4.5e-06, + "litellm_provider": "sambanova", + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_reasoning": true, + "source": "https://cloud.sambanova.ai/plans/pricing" + }, + "snowflake/claude-3-5-sonnet": { + "litellm_provider": "snowflake", + "max_input_tokens": 18000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "supports_computer_use": true + }, + "snowflake/deepseek-r1": { + "litellm_provider": "snowflake", + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "supports_reasoning": true + }, + "snowflake/gemma-7b": { + "litellm_provider": "snowflake", + "max_input_tokens": 8000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/jamba-1.5-large": { + "litellm_provider": "snowflake", + "max_input_tokens": 256000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/jamba-1.5-mini": { + "litellm_provider": "snowflake", + "max_input_tokens": 256000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/jamba-instruct": { + "litellm_provider": "snowflake", + "max_input_tokens": 256000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama2-70b-chat": { + "litellm_provider": "snowflake", + "max_input_tokens": 4096, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3-70b": { + "litellm_provider": "snowflake", + "max_input_tokens": 8000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3-8b": { + "litellm_provider": "snowflake", + "max_input_tokens": 8000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3.1-405b": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3.1-70b": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3.1-8b": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3.2-1b": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3.2-3b": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/llama3.3-70b": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/mistral-7b": { + "litellm_provider": "snowflake", + "max_input_tokens": 32000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/mistral-large": { + "litellm_provider": "snowflake", + "max_input_tokens": 32000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/mistral-large2": { + "litellm_provider": "snowflake", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/mixtral-8x7b": { + "litellm_provider": "snowflake", + "max_input_tokens": 32000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/reka-core": { + "litellm_provider": "snowflake", + "max_input_tokens": 32000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/reka-flash": { + "litellm_provider": "snowflake", + "max_input_tokens": 100000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/snowflake-arctic": { + "litellm_provider": "snowflake", + "max_input_tokens": 4096, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/snowflake-llama-3.1-405b": { + "litellm_provider": "snowflake", + "max_input_tokens": 8000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "snowflake/snowflake-llama-3.3-70b": { + "litellm_provider": "snowflake", + "max_input_tokens": 8000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat" + }, + "stability/sd3": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.065, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/sd3-large": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.065, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/sd3-large-turbo": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/sd3-medium": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.035, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/sd3.5-large": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.065, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/sd3.5-large-turbo": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/sd3.5-medium": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.035, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/stable-image-ultra": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.08, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability/inpaint": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/outpaint": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.004, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/erase": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/search-and-replace": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/search-and-recolor": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/remove-background": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/replace-background-and-relight": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.008, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/sketch": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/structure": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/style": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.005, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/style-transfer": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.008, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/fast": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.002, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/conservative": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.04, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/creative": { + "litellm_provider": "stability", + "mode": "image_edit", + "output_cost_per_image": 0.06, + "supported_endpoints": [ + "/v1/images/edits" + ] + }, + "stability/stable-image-core": { + "litellm_provider": "stability", + "mode": "image_generation", + "output_cost_per_image": 0.03, + "supported_endpoints": [ + "/v1/images/generations" + ] + }, + "stability.sd3-5-large-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.08 + }, + "stability.sd3-large-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.08 + }, + "stability.stable-image-core-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.04 + }, + "stability.stable-conservative-upscale-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.4 + }, + "stability.stable-creative-upscale-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.6 + }, + "stability.stable-fast-upscale-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.03 + }, + "stability.stable-outpaint-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.06 + }, + "stability.stable-image-control-sketch-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-control-structure-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-erase-object-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-inpaint-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-remove-background-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-search-recolor-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-search-replace-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-image-style-guide-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.07 + }, + "stability.stable-style-transfer-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "mode": "image_edit", + "output_cost_per_image": 0.08 + }, + "stability.stable-image-core-v1:1": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.04 + }, + "stability.stable-image-ultra-v1:0": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.14 + }, + "stability.stable-image-ultra-v1:1": { + "litellm_provider": "bedrock", + "max_input_tokens": 77, + "max_tokens": 77, + "mode": "image_generation", + "output_cost_per_image": 0.14 + }, + "standard/1024-x-1024/dall-e-3": { + "input_cost_per_pixel": 3.81469e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "standard/1024-x-1792/dall-e-3": { + "input_cost_per_pixel": 4.359e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "standard/1792-x-1024/dall-e-3": { + "input_cost_per_pixel": 4.359e-08, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_pixel": 0.0 + }, + "linkup/search": { + "input_cost_per_query": 0.00587, + "litellm_provider": "linkup", + "mode": "search" + }, + "linkup/search-deep": { + "input_cost_per_query": 0.05867, + "litellm_provider": "linkup", + "mode": "search" + }, + "tavily/search": { + "input_cost_per_query": 0.008, + "litellm_provider": "tavily", + "mode": "search" + }, + "tavily/search-advanced": { + "input_cost_per_query": 0.016, + "litellm_provider": "tavily", + "mode": "search" + }, + "text-completion-codestral/codestral-2405": { + "input_cost_per_token": 0.0, + "litellm_provider": "text-completion-codestral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "completion", + "output_cost_per_token": 0.0, + "source": "https://docs.mistral.ai/capabilities/code_generation/" + }, + "text-completion-codestral/codestral-latest": { + "input_cost_per_token": 0.0, + "litellm_provider": "text-completion-codestral", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "completion", + "output_cost_per_token": 0.0, + "source": "https://docs.mistral.ai/capabilities/code_generation/" + }, + "text-embedding-004": { + "deprecation_date": "2026-01-14", + "input_cost_per_character": 2.5e-08, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 768, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models" + }, + "text-embedding-005": { + "input_cost_per_character": 2.5e-08, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 768, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models" + }, + "text-embedding-3-large": { + "input_cost_per_token": 1.3e-07, + "input_cost_per_token_batches": 6.5e-08, + "litellm_provider": "openai", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_cost_per_token_batches": 0.0, + "output_vector_size": 3072 + }, + "text-embedding-3-small": { + "input_cost_per_token": 2e-08, + "input_cost_per_token_batches": 1e-08, + "litellm_provider": "openai", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_cost_per_token_batches": 0.0, + "output_vector_size": 1536 + }, + "text-embedding-ada-002": { + "input_cost_per_token": 1e-07, + "litellm_provider": "openai", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 1536 + }, + "text-embedding-ada-002-v2": { + "input_cost_per_token": 1e-07, + "input_cost_per_token_batches": 5e-08, + "litellm_provider": "openai", + "max_input_tokens": 8191, + "max_tokens": 8191, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_cost_per_token_batches": 0.0 + }, + "text-embedding-large-exp-03-07": { + "input_cost_per_character": 2.5e-08, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 8192, + "max_tokens": 8192, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 3072, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models" + }, + "text-embedding-preview-0409": { + "input_cost_per_token": 6.25e-09, + "input_cost_per_token_batch_requests": 5e-09, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 3072, + "max_tokens": 3072, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 768, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "text-moderation-007": { + "input_cost_per_token": 0.0, + "litellm_provider": "openai", + "max_input_tokens": 32768, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "moderation", + "output_cost_per_token": 0.0 + }, + "text-moderation-latest": { + "input_cost_per_token": 0.0, + "litellm_provider": "openai", + "max_input_tokens": 32768, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "moderation", + "output_cost_per_token": 0.0 + }, + "text-moderation-stable": { + "input_cost_per_token": 0.0, + "litellm_provider": "openai", + "max_input_tokens": 32768, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "moderation", + "output_cost_per_token": 0.0 + }, + "text-multilingual-embedding-002": { + "input_cost_per_character": 2.5e-08, + "input_cost_per_token": 1e-07, + "litellm_provider": "vertex_ai-embedding-models", + "max_input_tokens": 2048, + "max_tokens": 2048, + "mode": "embedding", + "output_cost_per_token": 0, + "output_vector_size": 768, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models" + }, + "text-unicorn": { + "input_cost_per_token": 1e-05, + "litellm_provider": "vertex_ai-text-models", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "completion", + "output_cost_per_token": 2.8e-05, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "text-unicorn@001": { + "input_cost_per_token": 1e-05, + "litellm_provider": "vertex_ai-text-models", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "completion", + "output_cost_per_token": 2.8e-05, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models" + }, + "together-ai-21.1b-41b": { + "input_cost_per_token": 8e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 8e-07 + }, + "together-ai-4.1b-8b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 2e-07 + }, + "together-ai-41.1b-80b": { + "input_cost_per_token": 9e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 9e-07 + }, + "together-ai-8.1b-21b": { + "input_cost_per_token": 3e-07, + "litellm_provider": "together_ai", + "max_tokens": 1000, + "mode": "chat", + "output_cost_per_token": 3e-07 + }, + "together-ai-81.1b-110b": { + "input_cost_per_token": 1.8e-06, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 1.8e-06 + }, + "together-ai-embedding-151m-to-350m": { + "input_cost_per_token": 1.6e-08, + "litellm_provider": "together_ai", + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "together-ai-embedding-up-to-150m": { + "input_cost_per_token": 8e-09, + "litellm_provider": "together_ai", + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "together_ai/baai/bge-base-en-v1.5": { + "input_cost_per_token": 8e-09, + "litellm_provider": "together_ai", + "max_input_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 768 + }, + "together_ai/BAAI/bge-base-en-v1.5": { + "input_cost_per_token": 8e-09, + "litellm_provider": "together_ai", + "max_input_tokens": 512, + "mode": "embedding", + "output_cost_per_token": 0.0, + "output_vector_size": 768 + }, + "together-ai-up-to-4b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 1e-07 + }, + "together_ai/Qwen/Qwen2.5-72B-Instruct-Turbo": { + "litellm_provider": "together_ai", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen2.5-7B-Instruct-Turbo": { + "litellm_provider": "together_ai", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen3-235B-A22B-Instruct-2507-tput": { + "input_cost_per_token": 2e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 262000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://www.together.ai/models/qwen3-235b-a22b-instruct-2507-fp8", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen3-235B-A22B-Thinking-2507": { + "input_cost_per_token": 6.5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://www.together.ai/models/qwen3-235b-a22b-thinking-2507", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen3-235B-A22B-fp8-tput": { + "input_cost_per_token": 2e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 40000, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://www.together.ai/models/qwen3-235b-a22b-fp8-tput", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_tool_choice": false + }, + "together_ai/Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8": { + "input_cost_per_token": 2e-06, + "litellm_provider": "together_ai", + "max_input_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://www.together.ai/models/qwen3-coder-480b-a35b-instruct", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/deepseek-ai/DeepSeek-R1": { + "input_cost_per_token": 3e-06, + "litellm_provider": "together_ai", + "max_input_tokens": 128000, + "max_output_tokens": 20480, + "max_tokens": 20480, + "mode": "chat", + "output_cost_per_token": 7e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/deepseek-ai/DeepSeek-R1-0528-tput": { + "input_cost_per_token": 5.5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.19e-06, + "source": "https://www.together.ai/models/deepseek-r1-0528-throughput", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/deepseek-ai/DeepSeek-V3": { + "input_cost_per_token": 1.25e-06, + "litellm_provider": "together_ai", + "max_input_tokens": 65536, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/deepseek-ai/DeepSeek-V3.1": { + "input_cost_per_token": 6e-07, + "litellm_provider": "together_ai", + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.7e-06, + "source": "https://www.together.ai/models/deepseek-v3-1", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "max_input_tokens": 128000, + "max_output_tokens": 16384 + }, + "together_ai/meta-llama/Llama-3.2-3B-Instruct-Turbo": { + "litellm_provider": "together_ai", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo": { + "input_cost_per_token": 8.8e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 8.8e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Llama-3.3-70B-Instruct-Turbo-Free": { + "input_cost_per_token": 0, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 0, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8": { + "input_cost_per_token": 2.7e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 8.5e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Llama-4-Scout-17B-16E-Instruct": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 5.9e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo": { + "input_cost_per_token": 3.5e-06, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 3.5e-06, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo": { + "input_cost_per_token": 8.8e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 8.8e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 1.8e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/mistralai/Mistral-7B-Instruct-v0.1": { + "litellm_provider": "together_ai", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/mistralai/Mistral-Small-24B-Instruct-2501": { + "litellm_provider": "together_ai", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "together_ai/mistralai/Mixtral-8x7B-Instruct-v0.1": { + "input_cost_per_token": 6e-07, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/moonshotai/Kimi-K2-Instruct": { + "input_cost_per_token": 1e-06, + "litellm_provider": "together_ai", + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://www.together.ai/models/kimi-k2-instruct", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/openai/gpt-oss-120b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://www.together.ai/models/gpt-oss-120b", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/openai/gpt-oss-20b": { + "input_cost_per_token": 5e-08, + "litellm_provider": "together_ai", + "max_input_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2e-07, + "source": "https://www.together.ai/models/gpt-oss-20b", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/togethercomputer/CodeLlama-34b-Instruct": { + "litellm_provider": "together_ai", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "together_ai/zai-org/GLM-4.5-Air-FP8": { + "input_cost_per_token": 2e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.1e-06, + "source": "https://www.together.ai/models/glm-4-5-air", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/zai-org/GLM-4.6": { + "input_cost_per_token": 6e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "max_tokens": 200000, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "source": "https://www.together.ai/models/glm-4-6", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "together_ai/zai-org/GLM-4.7": { + "input_cost_per_token": 4.5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "max_tokens": 200000, + "mode": "chat", + "output_cost_per_token": 2e-06, + "source": "https://www.together.ai/models/glm-4-7", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "together_ai/moonshotai/Kimi-K2.5": { + "input_cost_per_token": 5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2.8e-06, + "source": "https://www.together.ai/models/kimi-k2-5", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_reasoning": true + }, + "together_ai/moonshotai/Kimi-K2-Instruct-0905": { + "input_cost_per_token": 1e-06, + "litellm_provider": "together_ai", + "max_input_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3e-06, + "source": "https://www.together.ai/models/kimi-k2-0905", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen3-Next-80B-A3B-Instruct": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://www.together.ai/models/qwen3-next-80b-a3b-instruct", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen3-Next-80B-A3B-Thinking": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://www.together.ai/models/qwen3-next-80b-a3b-thinking", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "together_ai/Qwen/Qwen3.5-397B-A17B": { + "input_cost_per_token": 6e-07, + "litellm_provider": "together_ai", + "max_input_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 3.6e-06, + "source": "https://www.together.ai/models/Qwen/Qwen3.5-397B-A17B", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "tts-1": { + "input_cost_per_character": 1.5e-05, + "litellm_provider": "openai", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "tts-1-hd": { + "input_cost_per_character": 3e-05, + "litellm_provider": "openai", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "aws_polly/standard": { + "input_cost_per_character": 4e-06, + "litellm_provider": "aws_polly", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ], + "source": "https://aws.amazon.com/polly/pricing/" + }, + "aws_polly/neural": { + "input_cost_per_character": 1.6e-05, + "litellm_provider": "aws_polly", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ], + "source": "https://aws.amazon.com/polly/pricing/" + }, + "aws_polly/long-form": { + "input_cost_per_character": 0.0001, + "litellm_provider": "aws_polly", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ], + "source": "https://aws.amazon.com/polly/pricing/" + }, + "aws_polly/generative": { + "input_cost_per_character": 3e-05, + "litellm_provider": "aws_polly", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ], + "source": "https://aws.amazon.com/polly/pricing/" + }, + "us.amazon.nova-lite-v1:0": { + "input_cost_per_token": 6e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 2.4e-07, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "us.amazon.nova-micro-v1:0": { + "input_cost_per_token": 3.5e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.4e-07, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true + }, + "us.amazon.nova-premier-v1:0": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 1.25e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": false, + "supports_response_schema": true, + "supports_vision": true + }, + "us.amazon.nova-pro-v1:0": { + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 300000, + "max_output_tokens": 10000, + "max_tokens": 10000, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_vision": true + }, + "us.anthropic.claude-3-5-haiku-20241022-v1:0": { + "cache_creation_input_token_cost": 1e-06, + "cache_read_input_token_cost": 8e-08, + "input_cost_per_token": 8e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "us.anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.375e-06, + "cache_creation_input_token_cost_above_1hr": 2.2e-06, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5.5e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "us.anthropic.claude-3-5-sonnet-20240620-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "us.anthropic.claude-3-5-sonnet-20241022-v2:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "us.anthropic.claude-3-7-sonnet-20250219-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "us.anthropic.claude-3-haiku-20240307-v1:0": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 2.5e-08, + "cache_creation_input_token_cost": 3.125e-07 + }, + "us.anthropic.claude-3-opus-20240229-v1:0": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_function_calling": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 1.5e-06, + "cache_creation_input_token_cost": 1.875e-05 + }, + "us.anthropic.claude-3-sonnet-20240229-v1:0": { + "input_cost_per_token": 3e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "cache_read_input_token_cost": 3e-07, + "cache_creation_input_token_cost": 3.75e-06 + }, + "us.anthropic.claude-opus-4-1-20250805-v1:0": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "us.anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_creation_input_token_cost_above_1hr": 6.6e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "input_cost_per_token_above_200k_tokens": 6.6e-06, + "output_cost_per_token_above_200k_tokens": 2.475e-05, + "cache_creation_input_token_cost_above_200k_tokens": 8.25e-06, + "cache_creation_input_token_cost_above_1hr_above_200k_tokens": 1.32e-05, + "cache_read_input_token_cost_above_200k_tokens": 6.6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "us-gov.anthropic.claude-sonnet-4-5-20250929-v1:0": { + "cache_creation_input_token_cost": 4.125e-06, + "cache_read_input_token_cost": 3.3e-07, + "input_cost_per_token": 3.3e-06, + "input_cost_per_token_above_200k_tokens": 6.6e-06, + "output_cost_per_token_above_200k_tokens": 2.475e-05, + "cache_creation_input_token_cost_above_200k_tokens": 8.25e-06, + "cache_read_input_token_cost_above_200k_tokens": 6.6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.65e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "au.anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.375e-06, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5.5e-06, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true + }, + "us.anthropic.claude-opus-4-20250514-v1:0": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "us.anthropic.claude-opus-4-5-20251101-v1:0": { + "cache_creation_input_token_cost": 6.875e-06, + "cache_creation_input_token_cost_above_1hr": 1.1e-05, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 5.5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.75e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_native_structured_output": true + }, + "global.anthropic.claude-opus-4-5-20251101-v1:0": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_creation_input_token_cost_above_1hr": 1e-05, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_native_structured_output": true + }, + "eu.anthropic.claude-opus-4-5-20251101-v1:0": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_native_structured_output": true + }, + "us.anthropic.claude-sonnet-4-20250514-v1:0": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "us.deepseek.r1-v1:0": { + "input_cost_per_token": 1.35e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 5.4e-06, + "supports_function_calling": false, + "supports_reasoning": true, + "supports_tool_choice": false + }, + "us.deepseek.v3.2": { + "input_cost_per_token": 6.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 1.85e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "eu.deepseek.v3.2": { + "input_cost_per_token": 7.4e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "mode": "chat", + "output_cost_per_token": 2.22e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "us.meta.llama3-1-405b-instruct-v1:0": { + "input_cost_per_token": 5.32e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.6e-05, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama3-1-70b-instruct-v1:0": { + "input_cost_per_token": 9.9e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 9.9e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama3-1-8b-instruct-v1:0": { + "input_cost_per_token": 2.2e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 2.2e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama3-2-11b-instruct-v1:0": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3.5e-07, + "supports_function_calling": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "us.meta.llama3-2-1b-instruct-v1:0": { + "input_cost_per_token": 1e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama3-2-3b-instruct-v1:0": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama3-2-90b-instruct-v1:0": { + "input_cost_per_token": 2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_tool_choice": false, + "supports_vision": true + }, + "us.meta.llama3-3-70b-instruct-v1:0": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama4-maverick-17b-instruct-v1:0": { + "input_cost_per_token": 2.4e-07, + "input_cost_per_token_batches": 1.2e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 9.7e-07, + "output_cost_per_token_batches": 4.85e-07, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.meta.llama4-scout-17b-instruct-v1:0": { + "input_cost_per_token": 1.7e-07, + "input_cost_per_token_batches": 8.5e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6.6e-07, + "output_cost_per_token_batches": 3.3e-07, + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": false + }, + "us.mistral.pixtral-large-2502-v1:0": { + "input_cost_per_token": 2e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": false + }, + "v0/v0-1.0-md": { + "input_cost_per_token": 3e-06, + "litellm_provider": "v0", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "v0/v0-1.5-lg": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "v0", + "max_input_tokens": 512000, + "max_output_tokens": 512000, + "max_tokens": 512000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "v0/v0-1.5-md": { + "input_cost_per_token": 3e-06, + "litellm_provider": "v0", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/alibaba/qwen-3-14b": { + "input_cost_per_token": 8e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 40960, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2.4e-07 + }, + "vercel_ai_gateway/alibaba/qwen-3-235b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 40960, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07 + }, + "vercel_ai_gateway/alibaba/qwen-3-30b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 40960, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 3e-07 + }, + "vercel_ai_gateway/alibaba/qwen-3-32b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 40960, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/alibaba/qwen3-coder": { + "input_cost_per_token": 4e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 262144, + "max_output_tokens": 66536, + "max_tokens": 66536, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/amazon/nova-lite": { + "input_cost_per_token": 6e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 300000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.4e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/amazon/nova-micro": { + "input_cost_per_token": 3.5e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.4e-07, + "supports_function_calling": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/amazon/nova-pro": { + "input_cost_per_token": 8e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 300000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/amazon/titan-embed-text-v2": { + "input_cost_per_token": 2e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/anthropic/claude-3-haiku": { + "cache_creation_input_token_cost": 3e-07, + "cache_read_input_token_cost": 3e-08, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/anthropic/claude-3-opus": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/anthropic/claude-3.5-haiku": { + "cache_creation_input_token_cost": 1e-06, + "cache_read_input_token_cost": 8e-08, + "input_cost_per_token": 8e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/anthropic/claude-3.5-sonnet": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/anthropic/claude-3.7-sonnet": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/anthropic/claude-4-opus": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/anthropic/claude-4-sonnet": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/anthropic/claude-3-5-sonnet": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-3-5-sonnet-20241022": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-3-7-sonnet": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-haiku-4.5": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-opus-4": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-opus-4.1": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-opus-4.5": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-opus-4.6": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_minimal_reasoning_effort": true + }, + "vercel_ai_gateway/anthropic/claude-sonnet-4": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/anthropic/claude-sonnet-4.5": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vercel_ai_gateway/cohere/command-a": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 256000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/cohere/command-r": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/cohere/command-r-plus": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/cohere/embed-v4.0": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/deepseek/deepseek-r1": { + "input_cost_per_token": 5.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.19e-06, + "supports_tool_choice": true + }, + "vercel_ai_gateway/deepseek/deepseek-r1-distill-llama-70b": { + "input_cost_per_token": 7.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 9.9e-07, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/deepseek/deepseek-v3": { + "input_cost_per_token": 9e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 9e-07, + "supports_tool_choice": true + }, + "vercel_ai_gateway/google/gemini-2.0-flash": { + "deprecation_date": "2026-06-01", + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/google/gemini-2.0-flash-lite": { + "deprecation_date": "2026-06-01", + "input_cost_per_token": 7.5e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/google/gemini-2.5-flash": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1000000, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/google/gemini-2.5-pro": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/google/gemini-embedding-001": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/google/gemma-2-9b": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/google/text-embedding-005": { + "input_cost_per_token": 2.5e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/google/text-multilingual-embedding-002": { + "input_cost_per_token": 2.5e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/inception/mercury-coder-small": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-06 + }, + "vercel_ai_gateway/meta/llama-3-70b": { + "input_cost_per_token": 5.9e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 7.9e-07, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-3-8b": { + "input_cost_per_token": 5e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 8e-08, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-3.1-70b": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-3.1-8b": { + "input_cost_per_token": 5e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131000, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 8e-08, + "supports_function_calling": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/meta/llama-3.2-11b": { + "input_cost_per_token": 1.6e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.6e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-3.2-1b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1e-07 + }, + "vercel_ai_gateway/meta/llama-3.2-3b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_function_calling": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/meta/llama-3.2-90b": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-3.3-70b": { + "input_cost_per_token": 7.2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 7.2e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-4-maverick": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_tool_choice": true + }, + "vercel_ai_gateway/meta/llama-4-scout": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/mistral/codestral": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 256000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 9e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/mistral/codestral-embed": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/mistral/devstral-small": { + "input_cost_per_token": 7e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.8e-07, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/mistral/magistral-medium": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/mistral/magistral-small": { + "input_cost_per_token": 5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true + }, + "vercel_ai_gateway/mistral/ministral-3b": { + "input_cost_per_token": 4e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 4e-08, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/mistral/ministral-8b": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/mistral/mistral-embed": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "chat", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/mistral/mistral-large": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/mistral/mistral-saba-24b": { + "input_cost_per_token": 7.9e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 7.9e-07 + }, + "vercel_ai_gateway/mistral/mistral-small": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 3e-07, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/mistral/mixtral-8x22b-instruct": { + "input_cost_per_token": 1.2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 65536, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "supports_function_calling": true + }, + "vercel_ai_gateway/mistral/pixtral-12b": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/mistral/pixtral-large": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/moonshotai/kimi-k2": { + "input_cost_per_token": 5.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/morph/morph-v3-fast": { + "input_cost_per_token": 8e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32768, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.2e-06 + }, + "vercel_ai_gateway/morph/morph-v3-large": { + "input_cost_per_token": 9e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32768, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1.9e-06 + }, + "vercel_ai_gateway/openai/gpt-3.5-turbo": { + "input_cost_per_token": 5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 16385, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/openai/gpt-3.5-turbo-instruct": { + "input_cost_per_token": 1.5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 8192, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 2e-06 + }, + "vercel_ai_gateway/openai/gpt-4-turbo": { + "input_cost_per_token": 1e-05, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 3e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/openai/gpt-4.1": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/gpt-4.1-mini": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 4e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.6e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/gpt-4.1-nano": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 2.5e-08, + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 1047576, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/gpt-4o": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 2.5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/gpt-4o-mini": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/o1": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 7.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 6e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/o3": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/o3-mini": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 5.5e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/o4-mini": { + "cache_creation_input_token_cost": 0.0, + "cache_read_input_token_cost": 2.75e-07, + "input_cost_per_token": 1.1e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 100000, + "max_tokens": 100000, + "mode": "chat", + "output_cost_per_token": 4.4e-06, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_response_schema": true + }, + "vercel_ai_gateway/openai/text-embedding-3-large": { + "input_cost_per_token": 1.3e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/openai/text-embedding-3-small": { + "input_cost_per_token": 2e-08, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/openai/text-embedding-ada-002": { + "input_cost_per_token": 1e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 0, + "max_output_tokens": 0, + "max_tokens": 0, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "vercel_ai_gateway/perplexity/sonar": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 127000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1e-06 + }, + "vercel_ai_gateway/perplexity/sonar-pro": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 200000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 1.5e-05 + }, + "vercel_ai_gateway/perplexity/sonar-reasoning": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 127000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 5e-06 + }, + "vercel_ai_gateway/perplexity/sonar-reasoning-pro": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 127000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "mode": "chat", + "output_cost_per_token": 8e-06 + }, + "vercel_ai_gateway/vercel/v0-1.0-md": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/vercel/v0-1.5-md": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/xai/grok-2": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 4000, + "max_tokens": 4000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/xai/grok-2-vision": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_vision": true, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/xai/grok-3": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/xai/grok-3-fast": { + "input_cost_per_token": 5e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "supports_function_calling": true + }, + "vercel_ai_gateway/xai/grok-3-mini": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/xai/grok-3-mini-fast": { + "input_cost_per_token": 6e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/xai/grok-4": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/zai/glm-4.5": { + "input_cost_per_token": 6e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/zai/glm-4.5-air": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vercel_ai_gateway", + "max_input_tokens": 128000, + "max_output_tokens": 96000, + "max_tokens": 96000, + "mode": "chat", + "output_cost_per_token": 1.1e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vercel_ai_gateway/zai/glm-4.6": { + "litellm_provider": "vercel_ai_gateway", + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 4.5e-07, + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "max_tokens": 200000, + "mode": "chat", + "output_cost_per_token": 1.8e-06, + "source": "https://vercel.com/ai-gateway/models/glm-4.6", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/chirp": { + "input_cost_per_character": 3e-05, + "litellm_provider": "vertex_ai", + "mode": "audio_speech", + "source": "https://cloud.google.com/text-to-speech/pricing", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "vertex_ai/claude-3-5-haiku": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_tool_choice": true + }, + "vertex_ai/claude-3-5-haiku@20241022": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_tool_choice": true + }, + "vertex_ai/claude-haiku-4-5": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/haiku-4-5", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_native_streaming": true, + "supports_vision": true + }, + "vertex_ai/claude-haiku-4-5@20251001": { + "cache_creation_input_token_cost": 1.25e-06, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/haiku-4-5", + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_native_streaming": true, + "supports_vision": true + }, + "vertex_ai/claude-3-5-sonnet": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-5-sonnet@20240620": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-7-sonnet@20250219": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "deprecation_date": "2026-05-11", + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "vertex_ai/claude-3-haiku": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-haiku@20240307": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.25e-06, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-opus": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-opus@20240229": { + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-sonnet": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-3-sonnet@20240229": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-opus-4": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "vertex_ai/claude-opus-4-1": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "input_cost_per_token_batches": 7.5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "output_cost_per_token_batches": 3.75e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-opus-4-1@20250805": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "input_cost_per_token_batches": 7.5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "output_cost_per_token_batches": 3.75e-05, + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-opus-4-5": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "vertex_ai/claude-opus-4-5@20251101": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159, + "supports_native_streaming": true + }, + "vertex_ai/claude-opus-4-6": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "vertex_ai/claude-opus-4-6@default": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "vertex_ai/claude-opus-4-7": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "vertex_ai/claude-opus-4-7@default": { + "cache_creation_input_token_cost": 6.25e-06, + "cache_read_input_token_cost": 5e-07, + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": false, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_xhigh_reasoning_effort": true, + "tool_use_system_prompt_tokens": 346, + "supports_max_reasoning_effort": true, + "supports_minimal_reasoning_effort": true + }, + "vertex_ai/claude-sonnet-4-5": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "input_cost_per_token_batches": 1.5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_batches": 7.5e-06, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/claude-sonnet-4-6": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_minimal_reasoning_effort": true + }, + "vertex_ai/claude-sonnet-4-5@20250929": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "input_cost_per_token_batches": 1.5e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_batches": 7.5e-06, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_native_streaming": true + }, + "vertex_ai/claude-opus-4@20250514": { + "cache_creation_input_token_cost": 1.875e-05, + "cache_read_input_token_cost": 1.5e-06, + "input_cost_per_token": 1.5e-05, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 200000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 7.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "vertex_ai/claude-sonnet-4": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "vertex_ai/claude-sonnet-4@20250514": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_200k_tokens": 6e-06, + "output_cost_per_token_above_200k_tokens": 2.25e-05, + "cache_creation_input_token_cost_above_200k_tokens": 7.5e-06, + "cache_read_input_token_cost_above_200k_tokens": 6e-07, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 159 + }, + "vertex_ai/mistralai/codestral-2@001": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 9e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/codestral-2": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 9e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/codestral-2@001": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 9e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistralai/codestral-2": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 9e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/codestral-2501": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/codestral@2405": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/codestral@latest": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 6e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/deepseek-ai/deepseek-v3.1-maas": { + "input_cost_per_token": 1.35e-06, + "litellm_provider": "vertex_ai-deepseek_models", + "max_input_tokens": 163840, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 5.4e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_regions": [ + "us-central1" + ], + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "vertex_ai/deepseek-ai/deepseek-v3.2-maas": { + "input_cost_per_token": 5.6e-07, + "input_cost_per_token_batches": 2.8e-07, + "litellm_provider": "vertex_ai-deepseek_models", + "max_input_tokens": 163840, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1.68e-06, + "output_cost_per_token_batches": 8.4e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_regions": [ + "global" + ], + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "vertex_ai/deepseek-ai/deepseek-r1-0528-maas": { + "input_cost_per_token": 1.35e-06, + "litellm_provider": "vertex_ai-deepseek_models", + "max_input_tokens": 65336, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 5.4e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_regions": [ + "us-central1" + ], + "supports_assistant_prefill": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "vertex_ai/gemini-2.5-flash-image": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "max_pdf_size_mb": 30, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "image_generation", + "output_cost_per_image": 0.039, + "output_cost_per_image_token": 3e-05, + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 100000, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-generation#edit-an-image", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": false, + "tpm": 8000000 + }, + "vertex_ai/gemini-3-pro-image-preview": { + "input_cost_per_image": 0.0011, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.134, + "output_cost_per_image_token": 0.00012, + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_batches": 6e-06, + "source": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-pro-image" + }, + "vertex_ai/gemini-3.1-flash-image-preview": { + "input_cost_per_image": 0.00056, + "input_cost_per_token": 5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.0672, + "output_cost_per_image_token": 6e-05, + "output_cost_per_token": 3e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models" + }, + "vertex_ai/gemini-3.1-flash-lite-preview": { + "cache_read_input_token_cost": 2.5e-08, + "cache_read_input_token_cost_per_audio_token": 5e-08, + "input_cost_per_audio_token": 5e-07, + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-language-models", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65536, + "max_pdf_size_mb": 30, + "max_tokens": 65536, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 1.5e-06, + "output_cost_per_token": 1.5e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#gemini-models", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_audio_output": false, + "supports_code_execution": true, + "supports_file_search": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "supports_native_streaming": true, + "search_context_cost_per_query": { + "search_context_size_low": 0.014, + "search_context_size_medium": 0.014, + "search_context_size_high": 0.014 + }, + "web_search_billing_unit": "per_query" + }, + "vertex_ai/deep-research-pro-preview-12-2025": { + "input_cost_per_image": 0.0011, + "input_cost_per_token": 2e-06, + "input_cost_per_token_batches": 1e-06, + "litellm_provider": "vertex_ai-language-models", + "max_input_tokens": 65536, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.134, + "output_cost_per_image_token": 0.00012, + "output_cost_per_token": 1.2e-05, + "output_cost_per_token_batches": 6e-06, + "source": "https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-pro-image" + }, + "vertex_ai/imagegeneration@006": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.02, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/imagen-3.0-fast-generate-001": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.02, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/imagen-3.0-generate-001": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/imagen-3.0-generate-002": { + "deprecation_date": "2025-11-10", + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/imagen-3.0-capability-001": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/image/edit-insert-objects" + }, + "vertex_ai/imagen-4.0-fast-generate-001": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.02, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/imagen-4.0-generate-001": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.04, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/imagen-4.0-ultra-generate-001": { + "litellm_provider": "vertex_ai-image-models", + "mode": "image_generation", + "output_cost_per_image": 0.06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing" + }, + "vertex_ai/jamba-1.5": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vertex_ai-ai21_models", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "vertex_ai/jamba-1.5-large": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vertex_ai-ai21_models", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_tool_choice": true + }, + "vertex_ai/jamba-1.5-large@001": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vertex_ai-ai21_models", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 8e-06, + "supports_tool_choice": true + }, + "vertex_ai/jamba-1.5-mini": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vertex_ai-ai21_models", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "vertex_ai/jamba-1.5-mini@001": { + "input_cost_per_token": 2e-07, + "litellm_provider": "vertex_ai-ai21_models", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_tool_choice": true + }, + "vertex_ai/meta/llama-3.1-405b-instruct-maas": { + "input_cost_per_token": 5e-06, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 1.6e-05, + "source": "https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas", + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/meta/llama-3.1-70b-instruct-maas": { + "input_cost_per_token": 0.0, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas", + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/meta/llama-3.1-8b-instruct-maas": { + "input_cost_per_token": 0.0, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "metadata": { + "notes": "VertexAI states that The Llama 3.1 API service for llama-3.1-70b-instruct-maas and llama-3.1-8b-instruct-maas are in public preview and at no cost." + }, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas", + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/meta/llama-3.2-90b-vision-instruct-maas": { + "input_cost_per_token": 0.0, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 128000, + "max_output_tokens": 2048, + "max_tokens": 2048, + "metadata": { + "notes": "VertexAI states that The Llama 3.2 API service is at no cost during public preview, and will be priced as per dollar-per-1M-tokens at GA." + }, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.2-90b-vision-instruct-maas", + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/meta/llama-4-maverick-17b-128e-instruct-maas": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 1000000, + "max_output_tokens": 1000000, + "max_tokens": 1000000, + "mode": "chat", + "output_cost_per_token": 1.15e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/meta/llama-4-maverick-17b-16e-instruct-maas": { + "input_cost_per_token": 3.5e-07, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 1000000, + "max_output_tokens": 1000000, + "max_tokens": 1000000, + "mode": "chat", + "output_cost_per_token": 1.15e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/meta/llama-4-scout-17b-128e-instruct-maas": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 10000000, + "max_output_tokens": 10000000, + "max_tokens": 10000000, + "mode": "chat", + "output_cost_per_token": 7e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/meta/llama-4-scout-17b-16e-instruct-maas": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 10000000, + "max_output_tokens": 10000000, + "max_tokens": 10000000, + "mode": "chat", + "output_cost_per_token": 7e-07, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "code" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/meta/llama3-405b-instruct-maas": { + "input_cost_per_token": 0.0, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supports_tool_choice": true + }, + "vertex_ai/meta/llama3-70b-instruct-maas": { + "input_cost_per_token": 0.0, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supports_tool_choice": true + }, + "vertex_ai/meta/llama3-8b-instruct-maas": { + "input_cost_per_token": 0.0, + "litellm_provider": "vertex_ai-llama_models", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 0.0, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supports_tool_choice": true + }, + "vertex_ai/minimaxai/minimax-m2-maas": { + "input_cost_per_token": 3e-07, + "litellm_provider": "vertex_ai-minimax_models", + "max_input_tokens": 196608, + "max_output_tokens": 196608, + "max_tokens": 196608, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/moonshotai/kimi-k2-thinking-maas": { + "input_cost_per_token": 6e-07, + "litellm_provider": "vertex_ai-moonshot_models", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "vertex_ai/zai-org/glm-4.7-maas": { + "input_cost_per_token": 6e-07, + "litellm_provider": "vertex_ai-zai_models", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#partner-models", + "supported_regions": [ + "global" + ], + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "vertex_ai/zai-org/glm-5-maas": { + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-zai_models", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing#glm-models", + "supported_regions": [ + "global" + ], + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-medium-3": { + "input_cost_per_token": 4e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-medium-3@001": { + "input_cost_per_token": 4e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistralai/mistral-medium-3": { + "input_cost_per_token": 4e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistralai/mistral-medium-3@001": { + "input_cost_per_token": 4e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 2e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-large-2411": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-large@2407": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-large@2411-001": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-large@latest": { + "input_cost_per_token": 2e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 6e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-nemo@2407": { + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-nemo@latest": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1.5e-07, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-small-2503": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true + }, + "vertex_ai/mistral-small-2503@001": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-mistral_models", + "max_input_tokens": 32000, + "max_output_tokens": 8191, + "max_tokens": 8191, + "mode": "chat", + "output_cost_per_token": 3e-06, + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/mistral-ocr-2505": { + "litellm_provider": "vertex_ai", + "mode": "ocr", + "ocr_cost_per_page": 0.0005, + "supported_endpoints": [ + "/v1/ocr" + ], + "source": "https://cloud.google.com/generative-ai-app-builder/pricing" + }, + "vertex_ai/deepseek-ai/deepseek-ocr-maas": { + "litellm_provider": "vertex_ai", + "mode": "ocr", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "ocr_cost_per_page": 0.0003, + "source": "https://cloud.google.com/vertex-ai/pricing", + "supported_regions": [ + "us-central1" + ] + }, + "vertex_ai/openai/gpt-oss-120b-maas": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai-openai_models", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 6e-07, + "source": "https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss-120b-maas", + "supports_reasoning": true + }, + "vertex_ai/openai/gpt-oss-20b-maas": { + "input_cost_per_token": 7.5e-08, + "litellm_provider": "vertex_ai-openai_models", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 3e-07, + "source": "https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss-120b-maas", + "supports_reasoning": true + }, + "vertex_ai/qwen/qwen3-235b-a22b-instruct-2507-maas": { + "input_cost_per_token": 2.5e-07, + "litellm_provider": "vertex_ai-qwen_models", + "max_input_tokens": 262144, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_regions": [ + "global", + "us-south1" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/qwen/qwen3-coder-480b-a35b-instruct-maas": { + "input_cost_per_token": 1e-06, + "litellm_provider": "vertex_ai-qwen_models", + "max_input_tokens": 262144, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_regions": [ + "global" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/qwen/qwen3-next-80b-a3b-instruct-maas": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai-qwen_models", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_regions": [ + "global" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/qwen/qwen3-next-80b-a3b-thinking-maas": { + "input_cost_per_token": 1.5e-07, + "litellm_provider": "vertex_ai-qwen_models", + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "mode": "chat", + "output_cost_per_token": 1.2e-06, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_regions": [ + "global" + ], + "supports_function_calling": true, + "supports_tool_choice": true + }, + "vertex_ai/veo-2.0-generate-001": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.35, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "vertex_ai/veo-3.0-fast-generate-001": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.15, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "vertex_ai/veo-3.0-generate-001": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.4, + "source": "https://ai.google.dev/gemini-api/docs/video", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "vertex_ai/veo-3.1-generate-preview": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.4, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "vertex_ai/veo-3.1-fast-generate-preview": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.15, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "vertex_ai/veo-3.1-generate-001": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.4, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "vertex_ai/veo-3.1-fast-generate-001": { + "litellm_provider": "vertex_ai-video-models", + "max_input_tokens": 1024, + "max_tokens": 1024, + "mode": "video_generation", + "output_cost_per_second": 0.15, + "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ] + }, + "voyage/rerank-2": { + "input_cost_per_token": 5e-08, + "litellm_provider": "voyage", + "max_input_tokens": 16000, + "max_output_tokens": 16000, + "max_query_tokens": 16000, + "max_tokens": 16000, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "voyage/rerank-2-lite": { + "input_cost_per_token": 2e-08, + "litellm_provider": "voyage", + "max_input_tokens": 8000, + "max_output_tokens": 8000, + "max_query_tokens": 8000, + "max_tokens": 8000, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "voyage/rerank-2.5": { + "input_cost_per_token": 5e-08, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_query_tokens": 32000, + "max_tokens": 32000, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "voyage/rerank-2.5-lite": { + "input_cost_per_token": 2e-08, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_query_tokens": 32000, + "max_tokens": 32000, + "mode": "rerank", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-2": { + "input_cost_per_token": 1e-07, + "litellm_provider": "voyage", + "max_input_tokens": 4000, + "max_tokens": 4000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-3": { + "input_cost_per_token": 6e-08, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-3-large": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-3-lite": { + "input_cost_per_token": 2e-08, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-3.5": { + "input_cost_per_token": 6e-08, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-3.5-lite": { + "input_cost_per_token": 2e-08, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-code-2": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "voyage", + "max_input_tokens": 16000, + "max_tokens": 16000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-code-3": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-context-3": { + "input_cost_per_token": 1.8e-07, + "litellm_provider": "voyage", + "max_input_tokens": 120000, + "max_tokens": 120000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-finance-2": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-large-2": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "voyage", + "max_input_tokens": 16000, + "max_tokens": 16000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-law-2": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "voyage", + "max_input_tokens": 16000, + "max_tokens": 16000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-lite-01": { + "input_cost_per_token": 1e-07, + "litellm_provider": "voyage", + "max_input_tokens": 4096, + "max_tokens": 4096, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-lite-02-instruct": { + "input_cost_per_token": 1e-07, + "litellm_provider": "voyage", + "max_input_tokens": 4000, + "max_tokens": 4000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "voyage/voyage-multimodal-3": { + "input_cost_per_token": 1.2e-07, + "litellm_provider": "voyage", + "max_input_tokens": 32000, + "max_tokens": 32000, + "mode": "embedding", + "output_cost_per_token": 0.0 + }, + "wandb/openai/gpt-oss-120b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 0.015, + "output_cost_per_token": 0.06, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/openai/gpt-oss-20b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 0.005, + "output_cost_per_token": 0.02, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/zai-org/GLM-4.5": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 0.055, + "output_cost_per_token": 0.2, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/Qwen/Qwen3-235B-A22B-Instruct-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 0.01, + "output_cost_per_token": 0.01, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/Qwen/Qwen3-Coder-480B-A35B-Instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 0.1, + "output_cost_per_token": 0.15, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/Qwen/Qwen3-235B-A22B-Thinking-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 0.01, + "output_cost_per_token": 0.01, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/moonshotai/Kimi-K2-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.5e-06, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/moonshotai/Kimi-K2.5": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "cache_read_input_token_cost": 1e-07, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 3e-06, + "litellm_provider": "wandb", + "mode": "chat", + "source": "https://wandb.ai/inference/coreweave/cw_moonshotai_Kimi-K2.5", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_vision": true + }, + "wandb/MiniMaxAI/MiniMax-M2.5": { + "max_tokens": 197000, + "max_input_tokens": 197000, + "max_output_tokens": 197000, + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "wandb", + "mode": "chat", + "source": "https://wandb.ai/inference/coreweave/cw_MiniMaxAI_MiniMax-M2.5", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true + }, + "wandb/meta-llama/Llama-3.1-8B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 0.022, + "output_cost_per_token": 0.022, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/deepseek-ai/DeepSeek-V3.1": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 0.055, + "output_cost_per_token": 0.165, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/deepseek-ai/DeepSeek-R1-0528": { + "max_tokens": 161000, + "max_input_tokens": 161000, + "max_output_tokens": 161000, + "input_cost_per_token": 0.135, + "output_cost_per_token": 0.54, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/deepseek-ai/DeepSeek-V3-0324": { + "max_tokens": 161000, + "max_input_tokens": 161000, + "max_output_tokens": 161000, + "input_cost_per_token": 0.114, + "output_cost_per_token": 0.275, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/meta-llama/Llama-3.3-70B-Instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 0.071, + "output_cost_per_token": 0.071, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/meta-llama/Llama-4-Scout-17B-16E-Instruct": { + "max_tokens": 64000, + "max_input_tokens": 64000, + "max_output_tokens": 64000, + "input_cost_per_token": 0.017, + "output_cost_per_token": 0.066, + "litellm_provider": "wandb", + "mode": "chat" + }, + "wandb/microsoft/Phi-4-mini-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 0.008, + "output_cost_per_token": 0.035, + "litellm_provider": "wandb", + "mode": "chat" + }, + "watsonx/ibm/granite-3-8b-instruct": { + "input_cost_per_token": 2e-07, + "litellm_provider": "watsonx", + "max_input_tokens": 8192, + "max_output_tokens": 1024, + "max_tokens": 1024, + "mode": "chat", + "output_cost_per_token": 2e-07, + "supports_audio_input": false, + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "watsonx/mistralai/mistral-large": { + "input_cost_per_token": 3e-06, + "litellm_provider": "watsonx", + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_audio_input": false, + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": false, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": false + }, + "watsonx/bigscience/mt0-xxl-13b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 0.0005, + "output_cost_per_token": 0.002, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/core42/jais-13b-chat": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 0.0005, + "output_cost_per_token": 0.002, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/google/flan-t5-xl-3b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-13b-chat-v2": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-13b-instruct-v2": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-3-3-8b-instruct": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/ibm/granite-4-h-small": { + "max_tokens": 20480, + "max_input_tokens": 20480, + "max_output_tokens": 20480, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 2.5e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/ibm/granite-guardian-3-2-2b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-guardian-3-3-8b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-ttm-1024-96-r2": { + "max_tokens": 512, + "max_input_tokens": 512, + "max_output_tokens": 512, + "input_cost_per_token": 3.8e-07, + "output_cost_per_token": 3.8e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-ttm-1536-96-r2": { + "max_tokens": 512, + "max_input_tokens": 512, + "max_output_tokens": 512, + "input_cost_per_token": 3.8e-07, + "output_cost_per_token": 3.8e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-ttm-512-96-r2": { + "max_tokens": 512, + "max_input_tokens": 512, + "max_output_tokens": 512, + "input_cost_per_token": 3.8e-07, + "output_cost_per_token": 3.8e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/ibm/granite-vision-3-2-2b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": true + }, + "watsonx/meta-llama/llama-3-2-11b-vision-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 3.5e-07, + "output_cost_per_token": 3.5e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "watsonx/meta-llama/llama-3-2-1b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/meta-llama/llama-3-2-3b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 1.5e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/meta-llama/llama-3-2-90b-vision-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-06, + "output_cost_per_token": 2e-06, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": true + }, + "watsonx/meta-llama/llama-3-3-70b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 7.1e-07, + "output_cost_per_token": 7.1e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/meta-llama/llama-4-maverick-17b": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 3.5e-07, + "output_cost_per_token": 1.4e-06, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/meta-llama/llama-guard-3-11b-vision": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 3.5e-07, + "output_cost_per_token": 3.5e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": true + }, + "watsonx/mistralai/mistral-medium-2505": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 3e-06, + "output_cost_per_token": 1e-05, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/mistralai/mistral-small-2503": { + "max_tokens": 32000, + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 3e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/mistralai/mistral-small-3-1-24b-instruct-2503": { + "max_tokens": 32000, + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 3e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_vision": false + }, + "watsonx/mistralai/pixtral-12b-2409": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 3.5e-07, + "output_cost_per_token": 3.5e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": true + }, + "watsonx/openai/gpt-oss-120b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/sdaia/allam-1-13b-instruct": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 1.8e-06, + "output_cost_per_token": 1.8e-06, + "litellm_provider": "watsonx", + "mode": "chat", + "supports_function_calling": false, + "supports_parallel_function_calling": false, + "supports_vision": false + }, + "watsonx/whisper-large-v3-turbo": { + "input_cost_per_second": 0.0001, + "output_cost_per_second": 0.0001, + "litellm_provider": "watsonx", + "mode": "audio_transcription", + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "whisper-1": { + "input_cost_per_second": 0.0001, + "litellm_provider": "openai", + "mode": "audio_transcription", + "output_cost_per_second": 0.0001, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "xai/grok-2": { + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-2-1212": { + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-2-latest": { + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-2-vision": { + "input_cost_per_image": 2e-06, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-2-vision-1212": { + "deprecation_date": "2026-02-28", + "input_cost_per_image": 2e-06, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-2-vision-latest": { + "input_cost_per_image": 2e-06, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-3": { + "cache_read_input_token_cost": 7.5e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-beta": { + "cache_read_input_token_cost": 7.5e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-fast-beta": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-fast-latest": { + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 2.5e-05, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-latest": { + "cache_read_input_token_cost": 7.5e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-mini": { + "cache_read_input_token_cost": 7.5e-08, + "deprecation_date": "2026-02-28", + "input_cost_per_token": 3e-07, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-07, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-mini-beta": { + "cache_read_input_token_cost": 7.5e-08, + "deprecation_date": "2026-02-28", + "input_cost_per_token": 3e-07, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-07, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-mini-fast": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-mini-fast-beta": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-mini-fast-latest": { + "cache_read_input_token_cost": 1.5e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 4e-06, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-3-mini-latest": { + "cache_read_input_token_cost": 7.5e-08, + "input_cost_per_token": 3e-07, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 5e-07, + "source": "https://x.ai/api#pricing", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": false, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-4": { + "input_cost_per_token": 3e-06, + "litellm_provider": "xai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-4-fast-reasoning": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-4-fast-non-reasoning": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-4-0709": { + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_128k_tokens": 6e-06, + "litellm_provider": "xai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_128k_tokens": 3e-05, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-4-latest": { + "input_cost_per_token": 3e-06, + "input_cost_per_token_above_128k_tokens": 6e-06, + "litellm_provider": "xai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "output_cost_per_token_above_128k_tokens": 3e-05, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_web_search": true + }, + "xai/grok-4-1-fast": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models/grok-4-1-fast-reasoning", + "supports_audio_input": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4-1-fast-reasoning": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models/grok-4-1-fast-reasoning", + "supports_audio_input": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4-1-fast-reasoning-latest": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models/grok-4-1-fast-reasoning", + "supports_audio_input": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4-1-fast-non-reasoning": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models/grok-4-1-fast-non-reasoning", + "supports_audio_input": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4-1-fast-non-reasoning-latest": { + "cache_read_input_token_cost": 5e-08, + "input_cost_per_token": 2e-07, + "input_cost_per_token_above_128k_tokens": 4e-07, + "litellm_provider": "xai", + "max_input_tokens": 2000000.0, + "max_output_tokens": 2000000.0, + "max_tokens": 2000000.0, + "mode": "chat", + "output_cost_per_token": 5e-07, + "output_cost_per_token_above_128k_tokens": 1e-06, + "source": "https://docs.x.ai/docs/models/grok-4-1-fast-non-reasoning", + "supports_audio_input": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4.20-multi-agent-beta-0309": { + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 2000000, + "max_output_tokens": 2000000, + "max_tokens": 2000000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4.20-beta-0309-reasoning": { + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 2000000, + "max_output_tokens": 2000000, + "max_tokens": 2000000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4.20-0309-reasoning": { + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 2000000, + "max_output_tokens": 2000000, + "max_tokens": 2000000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-4.20-beta-0309-non-reasoning": { + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 2e-06, + "litellm_provider": "xai", + "max_input_tokens": 2000000, + "max_output_tokens": 2000000, + "max_tokens": 2000000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-beta": { + "input_cost_per_token": 5e-06, + "litellm_provider": "xai", + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "xai/grok-code-fast": { + "cache_read_input_token_cost": 2e-08, + "input_cost_per_token": 2e-07, + "litellm_provider": "xai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "xai/grok-code-fast-1": { + "cache_read_input_token_cost": 2e-08, + "input_cost_per_token": 2e-07, + "litellm_provider": "xai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "xai/grok-code-fast-1-0825": { + "cache_read_input_token_cost": 2e-08, + "input_cost_per_token": 2e-07, + "litellm_provider": "xai", + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "max_tokens": 256000, + "mode": "chat", + "output_cost_per_token": 1.5e-06, + "source": "https://docs.x.ai/docs/models", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true + }, + "xai/grok-vision-beta": { + "input_cost_per_image": 5e-06, + "input_cost_per_token": 5e-06, + "litellm_provider": "xai", + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true + }, + "zai.glm-4.7": { + "input_cost_per_token": 6e-07, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 2.2e-06, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "zai.glm-4.7-flash": { + "input_cost_per_token": 7e-08, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 4e-07, + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "zai.glm-5": { + "input_cost_per_token": 1e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 3.2e-06, + "source": "https://aws.amazon.com/bedrock/pricing/", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "zai/glm-5": { + "cache_creation_input_token_cost": 0, + "cache_read_input_token_cost": 2e-07, + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3.2e-06, + "litellm_provider": "zai", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-5-code": { + "cache_creation_input_token_cost": 0, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 5e-06, + "litellm_provider": "zai", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.7": { + "cache_creation_input_token_cost": 0, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.2e-06, + "litellm_provider": "zai", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.6": { + "cache_creation_input_token_cost": 0, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.2e-06, + "litellm_provider": "zai", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.5": { + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.2e-06, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.5v": { + "input_cost_per_token": 6e-07, + "output_cost_per_token": 1.8e-06, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.5-x": { + "input_cost_per_token": 2.2e-06, + "output_cost_per_token": 8.9e-06, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.5-air": { + "input_cost_per_token": 2e-07, + "output_cost_per_token": 1.1e-06, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.5-airx": { + "input_cost_per_token": 1.1e-06, + "output_cost_per_token": 4.5e-06, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4-32b-0414-128k": { + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "zai/glm-4.5-flash": { + "input_cost_per_token": 0, + "output_cost_per_token": 0, + "litellm_provider": "zai", + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "mode": "chat", + "supports_function_calling": true, + "supports_tool_choice": true, + "source": "https://docs.z.ai/guides/overview/pricing" + }, + "vertex_ai/search_api": { + "input_cost_per_query": 0.0015, + "litellm_provider": "vertex_ai", + "mode": "vector_store" + }, + "openai/container": { + "code_interpreter_cost_per_session": 0.03, + "litellm_provider": "openai", + "mode": "chat" + }, + "openai/sora-2": { + "litellm_provider": "openai", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.1, + "source": "https://platform.openai.com/docs/api-reference/videos", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "720x1280", + "1280x720" + ] + }, + "openai/sora-2-pro": { + "litellm_provider": "openai", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.3, + "source": "https://platform.openai.com/docs/api-reference/videos", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "720x1280", + "1280x720" + ] + }, + "openai/sora-2-pro-high-res": { + "litellm_provider": "openai", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.5, + "source": "https://platform.openai.com/docs/api-reference/videos", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "1024x1792", + "1792x1024" + ] + }, + "azure/sora-2": { + "litellm_provider": "azure", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.1, + "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "720x1280", + "1280x720" + ] + }, + "azure/sora-2-pro": { + "litellm_provider": "azure", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.3, + "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "720x1280", + "1280x720" + ] + }, + "azure/sora-2-pro-high-res": { + "litellm_provider": "azure", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.5, + "source": "https://azure.microsoft.com/en-us/products/ai-services/video-generation", + "supported_modalities": [ + "text" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "1024x1792", + "1792x1024" + ] + }, + "runwayml/gen4_turbo": { + "litellm_provider": "runwayml", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.05, + "source": "https://docs.dev.runwayml.com/guides/pricing/", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "1280x720", + "720x1280" + ], + "metadata": { + "comment": "5 credits per second @ $0.01 per credit = $0.05 per second" + } + }, + "runwayml/gen4_aleph": { + "litellm_provider": "runwayml", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.15, + "source": "https://docs.dev.runwayml.com/guides/pricing/", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "1280x720", + "720x1280" + ], + "metadata": { + "comment": "15 credits per second @ $0.01 per credit = $0.15 per second" + } + }, + "runwayml/gen3a_turbo": { + "litellm_provider": "runwayml", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.05, + "source": "https://docs.dev.runwayml.com/guides/pricing/", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "1280x720", + "720x1280" + ], + "metadata": { + "comment": "5 credits per second @ $0.01 per credit = $0.05 per second" + } + }, + "runwayml/gen4_image": { + "litellm_provider": "runwayml", + "mode": "image_generation", + "input_cost_per_image": 0.05, + "output_cost_per_image": 0.05, + "source": "https://docs.dev.runwayml.com/guides/pricing/", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "image" + ], + "supported_resolutions": [ + "1280x720", + "1920x1080" + ], + "metadata": { + "comment": "5 credits per 720p image or 8 credits per 1080p image @ $0.01 per credit. Using 5 credits ($0.05) as base cost" + } + }, + "runwayml/gen4_image_turbo": { + "litellm_provider": "runwayml", + "mode": "image_generation", + "input_cost_per_image": 0.02, + "output_cost_per_image": 0.02, + "source": "https://docs.dev.runwayml.com/guides/pricing/", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "image" + ], + "supported_resolutions": [ + "1280x720", + "1920x1080" + ], + "metadata": { + "comment": "2 credits per image (any resolution) @ $0.01 per credit = $0.02 per image" + } + }, + "runwayml/eleven_multilingual_v2": { + "litellm_provider": "runwayml", + "mode": "audio_speech", + "input_cost_per_character": 3e-07, + "source": "https://docs.dev.runwayml.com/guides/pricing/", + "metadata": { + "comment": "Estimated cost based on standard TTS pricing. RunwayML uses ElevenLabs models." + } + }, + "fireworks_ai/accounts/fireworks/models/qwen3-coder-480b-a35b-instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 4.5e-07, + "output_cost_per_token": 1.8e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat", + "supports_reasoning": true + }, + "fireworks_ai/accounts/fireworks/models/flux-kontext-pro": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 4e-08, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/SSD-1B": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1.3e-10, + "output_cost_per_token": 1.3e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/chronos-hermes-13b-v2": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-13b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-13b-instruct": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-13b-python": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-34b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-34b-instruct": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-34b-python": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-70b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-70b-instruct": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-70b-python": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-7b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-7b-instruct": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-llama-7b-python": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/code-qwen-1p5-7b": { + "max_tokens": 65536, + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/codegemma-2b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/codegemma-7b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/cogito-671b-v2-p1": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/cogito-v1-preview-llama-3b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/cogito-v1-preview-llama-70b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/cogito-v1-preview-llama-8b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/cogito-v1-preview-qwen-14b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/cogito-v1-preview-qwen-32b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/flux-kontext-max": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 8e-08, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/dbrx-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-1b-base": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-33b-instruct": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-7b-base": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-7b-base-v1p5": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-7b-instruct-v1p5": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-v2-lite-base": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-coder-v2-lite-instruct": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-prover-v2": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-0528-distill-qwen3-8b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-llama-70b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-llama-8b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-14b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-1p5b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-32b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-r1-distill-qwen-7b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v2-lite-chat": { + "max_tokens": 163840, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/deepseek-v2p5": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/devstral-small-2505": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/dobby-mini-unhinged-plus-llama-3-1-8b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/dobby-unhinged-llama-3-3-70b-new": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/dolphin-2-9-2-qwen2-72b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/dolphin-2p6-mixtral-8x7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/ernie-4p5-21b-a3b-pt": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/ernie-4p5-300b-a47b-pt": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/fare-20b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/firefunction-v1": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/firellava-13b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/firesearch-ocr-v6": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/fireworks-asr-large": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "audio_transcription" + }, + "fireworks_ai/accounts/fireworks/models/fireworks-asr-v2": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "audio_transcription" + }, + "fireworks_ai/accounts/fireworks/models/flux-1-dev": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/flux-1-dev-controlnet-union": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-09, + "output_cost_per_token": 1e-09, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/flux-1-dev-fp8": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 5e-10, + "output_cost_per_token": 5e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/flux-1-schnell": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/flux-1-schnell-fp8": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 3.5e-10, + "output_cost_per_token": 3.5e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/gemma-2b-it": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/gemma-3-27b-it": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/gemma-7b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/gemma-7b-it": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/gemma2-9b-it": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/glm-4p5v": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat", + "supports_reasoning": true + }, + "fireworks_ai/accounts/fireworks/models/gpt-oss-safeguard-120b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/gpt-oss-safeguard-20b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/hermes-2-pro-mistral-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/internvl3-38b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/internvl3-78b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/internvl3-8b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/japanese-stable-diffusion-xl": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1.3e-10, + "output_cost_per_token": 1.3e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/kat-coder": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/kat-dev-32b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/kat-dev-72b-exp": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-guard-2-8b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-guard-3-1b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-guard-3-8b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v2-13b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v2-13b-chat": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v2-70b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v2-70b-chat": { + "max_tokens": 2048, + "max_input_tokens": 2048, + "max_output_tokens": 2048, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v2-7b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v2-7b-chat": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3-70b-instruct": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3-70b-instruct-hf": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3-8b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3-8b-instruct-hf": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p1-405b-instruct-long": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p1-70b-instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p1-70b-instruct-1b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p1-nemotron-70b-instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p2-1b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p2-3b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llama-v3p3-70b-instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llamaguard-7b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/llava-yi-34b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/minimax-m1-80k": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/minimax-m2": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/ministral-3-14b-instruct-2512": { + "max_tokens": 256000, + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/ministral-3-3b-instruct-2512": { + "max_tokens": 256000, + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/ministral-3-8b-instruct-2512": { + "max_tokens": 256000, + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-7b-instruct-4k": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-7b-instruct-v0p2": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-7b-instruct-v3": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-7b-v0p2": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-large-3-fp8": { + "max_tokens": 256000, + "max_input_tokens": 256000, + "max_output_tokens": 256000, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-nemo-base-2407": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-nemo-instruct-2407": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mistral-small-24b-instruct-2501": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mixtral-8x22b": { + "max_tokens": 65536, + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mixtral-8x22b-instruct": { + "max_tokens": 65536, + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "input_cost_per_token": 1.2e-06, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mixtral-8x7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mixtral-8x7b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mixtral-8x7b-instruct-hf": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/mythomax-l2-13b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nemotron-nano-v2-12b-vl": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nous-capybara-7b-v1p9": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nous-hermes-2-mixtral-8x7b-dpo": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nous-hermes-2-yi-34b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nous-hermes-llama2-13b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nous-hermes-llama2-70b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nous-hermes-llama2-7b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nvidia-nemotron-nano-12b-v2": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/nvidia-nemotron-nano-9b-v2": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/openchat-3p5-0106-7b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/openhermes-2-mistral-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/openhermes-2p5-mistral-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/openorca-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/phi-2-3b": { + "max_tokens": 2048, + "max_input_tokens": 2048, + "max_output_tokens": 2048, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/phi-3-mini-128k-instruct": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/phi-3-vision-128k-instruct": { + "max_tokens": 32064, + "max_input_tokens": 32064, + "max_output_tokens": 32064, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/phind-code-llama-34b-python-v1": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/phind-code-llama-34b-v1": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/phind-code-llama-34b-v2": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/playground-v2-1024px-aesthetic": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1.3e-10, + "output_cost_per_token": 1.3e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/playground-v2-5-1024px-aesthetic": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1.3e-10, + "output_cost_per_token": 1.3e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/pythia-12b": { + "max_tokens": 2048, + "max_input_tokens": 2048, + "max_output_tokens": 2048, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen-qwq-32b-preview": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen-v2p5-14b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen-v2p5-7b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen1p5-72b-chat": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2-7b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2-vl-2b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2-vl-72b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2-vl-7b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-0p5b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-14b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-1p5b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-32b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-32b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-72b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-72b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-7b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-0p5b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-0p5b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-14b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-14b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-1p5b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-1p5b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct-128k": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct-32k-rope": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-32b-instruct-64k": { + "max_tokens": 65536, + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-3b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-3b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-coder-7b-instruct": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-math-72b-instruct": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-vl-32b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-vl-3b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-vl-72b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen2p5-vl-7b-instruct": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-0p6b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-14b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-1p7b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-1p7b-fp8-draft": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-1p7b-fp8-draft-131072": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-1p7b-fp8-draft-40960": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-235b-a22b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 2.2e-07, + "output_cost_per_token": 8.8e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-235b-a22b-instruct-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2.2e-07, + "output_cost_per_token": 8.8e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-235b-a22b-thinking-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2.2e-07, + "output_cost_per_token": 8.8e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b-instruct-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 5e-07, + "output_cost_per_token": 5e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-30b-a3b-thinking-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-32b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat", + "supports_reasoning": true + }, + "fireworks_ai/accounts/fireworks/models/qwen3-4b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-4b-instruct-2507": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-8b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat", + "supports_reasoning": true + }, + "fireworks_ai/accounts/fireworks/models/qwen3-coder-30b-a3b-instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-coder-480b-instruct-bf16": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-embedding-0p6b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "embedding" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-embedding-4b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "embedding" + }, + "fireworks_ai/accounts/fireworks/models/": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "embedding" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-next-80b-a3b-instruct": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-next-80b-a3b-thinking": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-reranker-0p6b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "rerank" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-reranker-4b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "rerank" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-reranker-8b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "max_output_tokens": 40960, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "rerank" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-vl-235b-a22b-instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2.2e-07, + "output_cost_per_token": 8.8e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-vl-235b-a22b-thinking": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 2.2e-07, + "output_cost_per_token": 8.8e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-vl-30b-a3b-instruct": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-vl-30b-a3b-thinking": { + "max_tokens": 262144, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-vl-32b-instruct": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwen3-vl-8b-instruct": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/qwq-32b": { + "max_tokens": 131072, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/rolm-ocr": { + "max_tokens": 128000, + "max_input_tokens": 128000, + "max_output_tokens": 128000, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/snorkel-mistral-7b-pairrm-dpo": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/stable-diffusion-xl-1024-v1-0": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1.3e-10, + "output_cost_per_token": 1.3e-10, + "litellm_provider": "fireworks_ai", + "mode": "image_generation" + }, + "fireworks_ai/accounts/fireworks/models/stablecode-3b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/starcoder-16b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/starcoder-7b": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/starcoder2-15b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/starcoder2-3b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/starcoder2-7b": { + "max_tokens": 16384, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/toppy-m-7b": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/whisper-v3": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "audio_transcription" + }, + "fireworks_ai/accounts/fireworks/models/whisper-v3-turbo": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 0.0, + "output_cost_per_token": 0.0, + "litellm_provider": "fireworks_ai", + "mode": "audio_transcription" + }, + "fireworks_ai/accounts/fireworks/models/yi-34b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/yi-34b-200k-capybara": { + "max_tokens": 200000, + "max_input_tokens": 200000, + "max_output_tokens": 200000, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/yi-34b-chat": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 9e-07, + "output_cost_per_token": 9e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/yi-6b": { + "max_tokens": 4096, + "max_input_tokens": 4096, + "max_output_tokens": 4096, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "fireworks_ai/accounts/fireworks/models/zephyr-7b-beta": { + "max_tokens": 32768, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "input_cost_per_token": 2e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "fireworks_ai", + "mode": "chat" + }, + "novita/deepseek/deepseek-v3.2": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.69e-07, + "output_cost_per_token": 4e-07, + "max_input_tokens": 163840, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.345e-07, + "input_cost_per_token_cache_hit": 1.345e-07, + "supports_reasoning": true + }, + "novita/minimax/minimax-m2.1": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "max_input_tokens": 204800, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 3e-08, + "input_cost_per_token_cache_hit": 3e-08 + }, + "novita/zai-org/glm-4.7": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.2e-06, + "max_input_tokens": 204800, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token_cache_hit": 1.1e-07, + "supports_reasoning": true + }, + "novita/xiaomimimo/mimo-v2-flash": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1e-07, + "output_cost_per_token": 3e-07, + "max_input_tokens": 262144, + "max_output_tokens": 32000, + "max_tokens": 32000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 2e-08, + "input_cost_per_token_cache_hit": 2e-08, + "supports_reasoning": true + }, + "novita/zai-org/autoglm-phone-9b-multilingual": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3.5e-08, + "output_cost_per_token": 1.38e-07, + "max_input_tokens": 65536, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_vision": true, + "supports_system_messages": true + }, + "novita/moonshotai/kimi-k2-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.5e-06, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/minimax/minimax-m2": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "max_input_tokens": 204800, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "cache_read_input_token_cost": 3e-08, + "input_cost_per_token_cache_hit": 3e-08, + "supports_reasoning": true + }, + "novita/paddlepaddle/paddleocr-vl": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2e-08, + "output_cost_per_token": 2e-08, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_vision": true, + "supports_system_messages": true + }, + "novita/deepseek/deepseek-v3.2-exp": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 4.1e-07, + "max_input_tokens": 163840, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/qwen/qwen3-vl-235b-a22b-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 9.8e-07, + "output_cost_per_token": 3.95e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_vision": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/zai-org/glm-4.6v": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 9e-07, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 5.5e-08, + "input_cost_per_token_cache_hit": 5.5e-08, + "supports_reasoning": true + }, + "novita/zai-org/glm-4.6": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5.5e-07, + "output_cost_per_token": 2.2e-06, + "max_input_tokens": 204800, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token_cache_hit": 1.1e-07, + "supports_reasoning": true + }, + "novita/kwaipilot/kat-coder-pro": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "max_input_tokens": 256000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 6e-08, + "input_cost_per_token_cache_hit": 6e-08 + }, + "novita/qwen/qwen3-next-80b-a3b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 1.5e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-next-80b-a3b-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 1.5e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/deepseek/deepseek-ocr": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-08, + "output_cost_per_token": 3e-08, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/deepseek/deepseek-v3.1-terminus": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.35e-07, + "input_cost_per_token_cache_hit": 1.35e-07, + "supports_reasoning": true + }, + "novita/qwen/qwen3-vl-235b-a22b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.5e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-max": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.11e-06, + "output_cost_per_token": 8.45e-06, + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/skywork/r1v4-lite": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2e-07, + "output_cost_per_token": 6e-07, + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/deepseek/deepseek-v3.1": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.35e-07, + "input_cost_per_token_cache_hit": 1.35e-07, + "supports_reasoning": true + }, + "novita/moonshotai/kimi-k2-0905": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.5e-06, + "max_input_tokens": 262144, + "max_output_tokens": 262144, + "max_tokens": 262144, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-coder-480b-a35b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.3e-06, + "max_input_tokens": 262144, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-coder-30b-a3b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 2.7e-07, + "max_input_tokens": 160000, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/openai/gpt-oss-120b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5e-08, + "output_cost_per_token": 2.5e-07, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/moonshotai/kimi-k2-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5.7e-07, + "output_cost_per_token": 2.3e-06, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/deepseek/deepseek-v3-0324": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 1.12e-06, + "max_input_tokens": 163840, + "max_output_tokens": 163840, + "max_tokens": 163840, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.35e-07, + "input_cost_per_token_cache_hit": 1.35e-07 + }, + "novita/zai-org/glm-4.5": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6e-07, + "output_cost_per_token": 2.2e-06, + "max_input_tokens": 131072, + "max_output_tokens": 98304, + "max_tokens": 98304, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token_cache_hit": 1.1e-07, + "supports_reasoning": true + }, + "novita/qwen/qwen3-235b-a22b-thinking-2507": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 3e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/meta-llama/llama-3.1-8b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2e-08, + "output_cost_per_token": 5e-08, + "max_input_tokens": 16384, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_system_messages": true + }, + "novita/google/gemma-3-12b-it": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5e-08, + "output_cost_per_token": 1e-07, + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/zai-org/glm-4.5v": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6e-07, + "output_cost_per_token": 1.8e-06, + "max_input_tokens": 65536, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 1.1e-07, + "input_cost_per_token_cache_hit": 1.1e-07, + "supports_reasoning": true + }, + "novita/openai/gpt-oss-20b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1.5e-07, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/qwen/qwen3-235b-a22b-instruct-2507": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 9e-08, + "output_cost_per_token": 5.8e-07, + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/deepseek/deepseek-r1-distill-qwen-14b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 1.5e-07, + "max_input_tokens": 32768, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/meta-llama/llama-3.3-70b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.35e-07, + "output_cost_per_token": 4e-07, + "max_input_tokens": 131072, + "max_output_tokens": 120000, + "max_tokens": 120000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/qwen/qwen-2.5-72b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3.8e-07, + "output_cost_per_token": 4e-07, + "max_input_tokens": 32000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/mistralai/mistral-nemo": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1.7e-07, + "max_input_tokens": 60288, + "max_output_tokens": 16000, + "max_tokens": 16000, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/minimaxai/minimax-m1-80k": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5.5e-07, + "output_cost_per_token": 2.2e-06, + "max_input_tokens": 1000000, + "max_output_tokens": 40000, + "max_tokens": 40000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/deepseek/deepseek-r1-0528": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-07, + "output_cost_per_token": 2.5e-06, + "max_input_tokens": 163840, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "cache_read_input_token_cost": 3.5e-07, + "input_cost_per_token_cache_hit": 3.5e-07, + "supports_reasoning": true + }, + "novita/deepseek/deepseek-r1-distill-qwen-32b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-07, + "output_cost_per_token": 3e-07, + "max_input_tokens": 64000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/meta-llama/llama-3-8b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 4e-08, + "output_cost_per_token": 4e-08, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_system_messages": true + }, + "novita/microsoft/wizardlm-2-8x22b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6.2e-07, + "output_cost_per_token": 6.2e-07, + "max_input_tokens": 65535, + "max_output_tokens": 8000, + "max_tokens": 8000, + "supports_system_messages": true + }, + "novita/deepseek/deepseek-r1-0528-qwen3-8b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 6e-08, + "output_cost_per_token": 9e-08, + "max_input_tokens": 128000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/deepseek/deepseek-r1-distill-llama-70b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 8e-07, + "output_cost_per_token": 8e-07, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/meta-llama/llama-3-70b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5.1e-07, + "output_cost_per_token": 7.4e-07, + "max_input_tokens": 8192, + "max_output_tokens": 8000, + "max_tokens": 8000, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-235b-a22b-fp8": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2e-07, + "output_cost_per_token": 8e-07, + "max_input_tokens": 40960, + "max_output_tokens": 20000, + "max_tokens": 20000, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/meta-llama/llama-4-maverick-17b-128e-instruct-fp8": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.7e-07, + "output_cost_per_token": 8.5e-07, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_vision": true, + "supports_system_messages": true + }, + "novita/meta-llama/llama-4-scout-17b-16e-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.8e-07, + "output_cost_per_token": 5.9e-07, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_vision": true, + "supports_system_messages": true + }, + "novita/nousresearch/hermes-2-pro-llama-3-8b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.4e-07, + "output_cost_per_token": 1.4e-07, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen2.5-vl-72b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 8e-07, + "output_cost_per_token": 8e-07, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_vision": true, + "supports_system_messages": true + }, + "novita/sao10k/l3-70b-euryale-v2.1": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.48e-06, + "output_cost_per_token": 1.48e-06, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/baidu/ernie-4.5-21B-a3b-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 2.8e-07, + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/sao10k/l3-8b-lunaris": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5e-08, + "output_cost_per_token": 5e-08, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/baichuan/baichuan-m2-32b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 7e-08, + "max_input_tokens": 131072, + "max_output_tokens": 131072, + "max_tokens": 131072, + "supports_system_messages": true + }, + "novita/baidu/ernie-4.5-vl-424b-a47b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 4.2e-07, + "output_cost_per_token": 1.25e-06, + "max_input_tokens": 123000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "supports_vision": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/baidu/ernie-4.5-300b-a47b-paddle": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.8e-07, + "output_cost_per_token": 1.1e-06, + "max_input_tokens": 123000, + "max_output_tokens": 12000, + "max_tokens": 12000, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/deepseek/deepseek-prover-v2-671b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-07, + "output_cost_per_token": 2.5e-06, + "max_input_tokens": 160000, + "max_output_tokens": 160000, + "max_tokens": 160000, + "supports_system_messages": true + }, + "novita/qwen/qwen3-32b-fp8": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1e-07, + "output_cost_per_token": 4.5e-07, + "max_input_tokens": 40960, + "max_output_tokens": 20000, + "max_tokens": 20000, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/qwen/qwen3-30b-a3b-fp8": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 9e-08, + "output_cost_per_token": 4.5e-07, + "max_input_tokens": 40960, + "max_output_tokens": 20000, + "max_tokens": 20000, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/google/gemma-3-27b-it": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.19e-07, + "output_cost_per_token": 2e-07, + "max_input_tokens": 98304, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_vision": true, + "supports_system_messages": true + }, + "novita/deepseek/deepseek-v3-turbo": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 4e-07, + "output_cost_per_token": 1.3e-06, + "max_input_tokens": 64000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/deepseek/deepseek-r1-turbo": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-07, + "output_cost_per_token": 2.5e-06, + "max_input_tokens": 64000, + "max_output_tokens": 16000, + "max_tokens": 16000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/Sao10K/L3-8B-Stheno-v3.2": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 5e-08, + "output_cost_per_token": 5e-08, + "max_input_tokens": 8192, + "max_output_tokens": 32000, + "max_tokens": 32000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/gryphe/mythomax-l2-13b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 9e-08, + "output_cost_per_token": 9e-08, + "max_input_tokens": 4096, + "max_output_tokens": 3200, + "max_tokens": 3200, + "supports_system_messages": true + }, + "novita/baidu/ernie-4.5-vl-28b-a3b-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3.9e-07, + "output_cost_per_token": 3.9e-07, + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "novita/qwen/qwen3-vl-8b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 8e-08, + "output_cost_per_token": 5e-07, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/zai-org/glm-4.5-air": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 8.5e-07, + "max_input_tokens": 131072, + "max_output_tokens": 98304, + "max_tokens": 98304, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/qwen/qwen3-vl-30b-a3b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2e-07, + "output_cost_per_token": 7e-07, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-vl-30b-a3b-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2e-07, + "output_cost_per_token": 1e-06, + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/qwen/qwen3-omni-30b-a3b-thinking": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 9.7e-07, + "max_input_tokens": 65536, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_reasoning": true, + "supports_audio_input": true + }, + "novita/qwen/qwen3-omni-30b-a3b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 9.7e-07, + "max_input_tokens": 65536, + "max_output_tokens": 16384, + "max_tokens": 16384, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_response_schema": true, + "supports_audio_input": true, + "supports_audio_output": true + }, + "novita/qwen/qwen-mt-plus": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 2.5e-07, + "output_cost_per_token": 7.5e-07, + "max_input_tokens": 16384, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_system_messages": true + }, + "novita/baidu/ernie-4.5-vl-28b-a3b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.4e-07, + "output_cost_per_token": 5.6e-07, + "max_input_tokens": 30000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/baidu/ernie-4.5-21B-a3b": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 2.8e-07, + "max_input_tokens": 120000, + "max_output_tokens": 8000, + "max_tokens": 8000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/qwen/qwen3-8b-fp8": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3.5e-08, + "output_cost_per_token": 1.38e-07, + "max_input_tokens": 128000, + "max_output_tokens": 20000, + "max_tokens": 20000, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/qwen/qwen3-4b-fp8": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-08, + "output_cost_per_token": 3e-08, + "max_input_tokens": 128000, + "max_output_tokens": 20000, + "max_tokens": 20000, + "supports_system_messages": true, + "supports_reasoning": true + }, + "novita/qwen/qwen2.5-7b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 7e-08, + "max_input_tokens": 32000, + "max_output_tokens": 32000, + "max_tokens": 32000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true, + "supports_response_schema": true + }, + "novita/meta-llama/llama-3.2-3b-instruct": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 3e-08, + "output_cost_per_token": 5e-08, + "max_input_tokens": 32768, + "max_output_tokens": 32000, + "max_tokens": 32000, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/sao10k/l31-70b-euryale-v2.2": { + "litellm_provider": "novita", + "mode": "chat", + "input_cost_per_token": 1.48e-06, + "output_cost_per_token": 1.48e-06, + "max_input_tokens": 8192, + "max_output_tokens": 8192, + "max_tokens": 8192, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_tool_choice": true, + "supports_system_messages": true + }, + "novita/qwen/qwen3-embedding-0.6b": { + "litellm_provider": "novita", + "mode": "embedding", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 0, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768 + }, + "novita/qwen/qwen3-embedding-8b": { + "litellm_provider": "novita", + "mode": "embedding", + "input_cost_per_token": 7e-08, + "output_cost_per_token": 0, + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096 + }, + "novita/baai/bge-m3": { + "litellm_provider": "novita", + "mode": "embedding", + "input_cost_per_token": 1e-08, + "output_cost_per_token": 1e-08, + "max_input_tokens": 8192, + "max_output_tokens": 96000, + "max_tokens": 96000 + }, + "novita/qwen/qwen3-reranker-8b": { + "litellm_provider": "novita", + "mode": "rerank", + "input_cost_per_token": 5e-08, + "output_cost_per_token": 5e-08, + "max_input_tokens": 32768, + "max_output_tokens": 4096, + "max_tokens": 4096 + }, + "novita/baai/bge-reranker-v2-m3": { + "litellm_provider": "novita", + "mode": "rerank", + "input_cost_per_token": 1e-08, + "output_cost_per_token": 1e-08, + "max_input_tokens": 8000, + "max_output_tokens": 8000, + "max_tokens": 8000 + }, + "llamagate/llama-3.1-8b": { + "max_tokens": 8192, + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "input_cost_per_token": 3e-08, + "output_cost_per_token": 5e-08, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/llama-3.2-3b": { + "max_tokens": 8192, + "max_input_tokens": 131072, + "max_output_tokens": 8192, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 8e-08, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/mistral-7b-v0.3": { + "max_tokens": 8192, + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 1.5e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/qwen3-8b": { + "max_tokens": 8192, + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "input_cost_per_token": 4e-08, + "output_cost_per_token": 1.4e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/dolphin3-8b": { + "max_tokens": 8192, + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 1.5e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/deepseek-r1-8b": { + "max_tokens": 16384, + "max_input_tokens": 65536, + "max_output_tokens": 16384, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "llamagate/deepseek-r1-7b-qwen": { + "max_tokens": 16384, + "max_input_tokens": 131072, + "max_output_tokens": 16384, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 1.5e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "llamagate/openthinker-7b": { + "max_tokens": 8192, + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "input_cost_per_token": 8e-08, + "output_cost_per_token": 1.5e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_reasoning": true + }, + "llamagate/qwen2.5-coder-7b": { + "max_tokens": 8192, + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 1.2e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/deepseek-coder-6.7b": { + "max_tokens": 4096, + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 1.2e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/codellama-7b": { + "max_tokens": 4096, + "max_input_tokens": 16384, + "max_output_tokens": 4096, + "input_cost_per_token": 6e-08, + "output_cost_per_token": 1.2e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true + }, + "llamagate/qwen3-vl-8b": { + "max_tokens": 8192, + "max_input_tokens": 32768, + "max_output_tokens": 8192, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 5.5e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "llamagate/llava-7b": { + "max_tokens": 2048, + "max_input_tokens": 4096, + "max_output_tokens": 2048, + "input_cost_per_token": 1e-07, + "output_cost_per_token": 2e-07, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_response_schema": true, + "supports_vision": true + }, + "llamagate/gemma3-4b": { + "max_tokens": 8192, + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "input_cost_per_token": 3e-08, + "output_cost_per_token": 8e-08, + "litellm_provider": "llamagate", + "mode": "chat", + "supports_function_calling": true, + "supports_response_schema": true, + "supports_vision": true + }, + "llamagate/nomic-embed-text": { + "max_tokens": 8192, + "max_input_tokens": 8192, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 0, + "litellm_provider": "llamagate", + "mode": "embedding" + }, + "llamagate/qwen3-embedding-8b": { + "max_tokens": 40960, + "max_input_tokens": 40960, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 0, + "litellm_provider": "llamagate", + "mode": "embedding" + }, + "sarvam/sarvam-m": { + "cache_creation_input_token_cost": 0, + "cache_creation_input_token_cost_above_1hr": 0, + "cache_read_input_token_cost": 0, + "input_cost_per_token": 0, + "litellm_provider": "sarvam", + "max_input_tokens": 8192, + "max_output_tokens": 32000, + "max_tokens": 32000, + "mode": "chat", + "output_cost_per_token": 0, + "supports_reasoning": true + }, + "tts-1-1106": { + "input_cost_per_character": 1.5e-05, + "litellm_provider": "openai", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "tts-1-hd-1106": { + "input_cost_per_character": 3e-05, + "litellm_provider": "openai", + "mode": "audio_speech", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "gpt-4o-mini-tts-2025-03-20": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "mode": "audio_speech", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_second": 0.00025, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/speech" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "audio" + ] + }, + "gpt-4o-mini-tts-2025-12-15": { + "input_cost_per_token": 2.5e-06, + "litellm_provider": "openai", + "mode": "audio_speech", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_second": 0.00025, + "output_cost_per_token": 1e-05, + "supported_endpoints": [ + "/v1/audio/speech" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "audio" + ] + }, + "gpt-4o-mini-transcribe-2025-03-20": { + "input_cost_per_audio_token": 3e-06, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 5e-06, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "gpt-4o-mini-transcribe-2025-12-15": { + "input_cost_per_audio_token": 3e-06, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 16000, + "max_output_tokens": 2000, + "mode": "audio_transcription", + "output_cost_per_token": 5e-06, + "supported_endpoints": [ + "/v1/audio/transcriptions" + ] + }, + "gpt-5-search-api": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false, + "supports_minimal_reasoning_effort": true + }, + "gpt-5-search-api-2025-10-14": { + "cache_read_input_token_cost": 1.25e-07, + "input_cost_per_token": 1.25e-06, + "litellm_provider": "openai", + "max_input_tokens": 272000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "output_cost_per_token": 1e-05, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "supports_none_reasoning_effort": false, + "supports_xhigh_reasoning_effort": false + }, + "gpt-realtime-mini-2025-10-06": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_audio_token_cost": 3e-07, + "cache_read_input_token_cost": 6e-08, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_image": 8e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "gpt-realtime-mini-2025-12-15": { + "cache_creation_input_audio_token_cost": 3e-07, + "cache_read_input_audio_token_cost": 3e-07, + "cache_read_input_token_cost": 6e-08, + "input_cost_per_audio_token": 1e-05, + "input_cost_per_image": 8e-07, + "input_cost_per_token": 6e-07, + "litellm_provider": "openai", + "max_input_tokens": 128000, + "max_output_tokens": 4096, + "max_tokens": 4096, + "mode": "chat", + "output_cost_per_audio_token": 2e-05, + "output_cost_per_token": 2.4e-06, + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true + }, + "sora-2": { + "litellm_provider": "openai", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.1, + "source": "https://platform.openai.com/docs/api-reference/videos", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "720x1280", + "1280x720" + ] + }, + "sora-2-pro": { + "litellm_provider": "openai", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.3, + "source": "https://platform.openai.com/docs/api-reference/videos", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "720x1280", + "1280x720" + ] + }, + "sora-2-pro-high-res": { + "litellm_provider": "openai", + "mode": "video_generation", + "output_cost_per_video_per_second": 0.5, + "source": "https://platform.openai.com/docs/api-reference/videos", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "video" + ], + "supported_resolutions": [ + "1024x1792", + "1792x1024" + ] + }, + "chatgpt-image-latest": { + "cache_read_input_image_token_cost": 2.5e-06, + "cache_read_input_token_cost": 1.25e-06, + "input_cost_per_image_token": 1e-05, + "input_cost_per_token": 5e-06, + "litellm_provider": "openai", + "mode": "image_generation", + "output_cost_per_image_token": 4e-05, + "supported_endpoints": [ + "/v1/images/generations", + "/v1/images/edits" + ] + }, + "gemini-2.0-flash-exp-image-generation": { + "input_cost_per_token": 0.0, + "litellm_provider": "gemini", + "max_images_per_prompt": 3000, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.039, + "output_cost_per_token": 0.0, + "source": "https://ai.google.dev/pricing", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_vision": true + }, + "gemini/gemini-2.0-flash-exp-image-generation": { + "input_cost_per_token": 0.0, + "litellm_provider": "gemini", + "max_images_per_prompt": 3000, + "max_input_tokens": 32768, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "image_generation", + "output_cost_per_image": 0.039, + "output_cost_per_token": 0.0, + "source": "https://ai.google.dev/pricing", + "supported_modalities": [ + "text", + "image" + ], + "supported_output_modalities": [ + "text", + "image" + ], + "supports_vision": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini/gemini-2.0-flash-lite-001": { + "cache_read_input_token_cost": 1.875e-08, + "deprecation_date": "2026-06-01", + "input_cost_per_audio_token": 7.5e-08, + "input_cost_per_token": 7.5e-08, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_pdf_size_mb": 50, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 3e-07, + "rpm": 4000, + "source": "https://ai.google.dev/gemini-api/docs/pricing#gemini-2.0-flash-lite", + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 4000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-2.5-flash-native-audio-latest": { + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true + }, + "gemini-2.5-flash-native-audio-preview-09-2025": { + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true + }, + "gemini-2.5-flash-native-audio-preview-12-2025": { + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true + }, + "gemini-3.1-flash-live-preview": { + "input_cost_per_audio_token": 3e-06, + "input_cost_per_image_token": 1e-06, + "input_cost_per_token": 7.5e-07, + "input_cost_per_video_per_second": 3.3333333333333335e-05, + "litellm_provider": "gemini", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_token": 4.5e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_vision": true, + "supports_web_search": true + }, + "gemini/gemini-2.5-flash-native-audio-latest": { + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini/gemini-2.5-flash-native-audio-preview-09-2025": { + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini/gemini-2.5-flash-native-audio-preview-12-2025": { + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_input_tokens": 1048576, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "audio" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini/gemini-3.1-flash-live-preview": { + "input_cost_per_audio_token": 3e-06, + "input_cost_per_image_token": 1e-06, + "input_cost_per_token": 7.5e-07, + "input_cost_per_video_per_second": 3.3333333333333335e-05, + "litellm_provider": "gemini", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "output_cost_per_audio_token": 1.2e-05, + "output_cost_per_token": 4.5e-06, + "source": "https://ai.google.dev/gemini-api/docs/pricing", + "supported_endpoints": [ + "/v1/realtime" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text", + "audio" + ], + "supports_audio_input": true, + "supports_audio_output": true, + "supports_function_calling": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "rpm": 10 + }, + "gemini-2.5-flash-preview-tts": { + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "mode": "audio_speech", + "output_cost_per_token": 2.5e-06, + "source": "https://ai.google.dev/pricing", + "supported_endpoints": [ + "/v1/audio/speech" + ] + }, + "gemini-flash-latest": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 100000, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 8000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-flash-lite-latest": { + "cache_read_input_token_cost": 1e-08, + "input_cost_per_audio_token": 3e-07, + "input_cost_per_token": 1e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 4e-07, + "output_cost_per_token": 4e-07, + "rpm": 15, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-lite", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 250000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-pro-latest": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "rpm": 2000, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 800000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini/gemini-pro-latest": { + "cache_read_input_token_cost": 1.25e-07, + "cache_read_input_token_cost_above_200k_tokens": 2.5e-07, + "input_cost_per_token": 1.25e-06, + "input_cost_per_token_above_200k_tokens": 2.5e-06, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_token": 1e-05, + "output_cost_per_token_above_200k_tokens": 1.5e-05, + "rpm": 2000, + "source": "https://cloud.google.com/vertex-ai/generative-ai/pricing", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_input": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_video_input": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 800000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "gemini-exp-1206": { + "cache_read_input_token_cost": 3e-08, + "input_cost_per_audio_token": 1e-06, + "input_cost_per_token": 3e-07, + "litellm_provider": "gemini", + "max_audio_length_hours": 8.4, + "max_audio_per_prompt": 1, + "max_images_per_prompt": 3000, + "max_input_tokens": 1048576, + "max_output_tokens": 65535, + "max_pdf_size_mb": 30, + "max_tokens": 65535, + "max_video_length": 1, + "max_videos_per_prompt": 10, + "mode": "chat", + "output_cost_per_reasoning_token": 2.5e-06, + "output_cost_per_token": 2.5e-06, + "rpm": 100000, + "source": "https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview", + "supported_endpoints": [ + "/v1/chat/completions", + "/v1/completions", + "/v1/batch" + ], + "supported_modalities": [ + "text", + "image", + "audio", + "video" + ], + "supported_output_modalities": [ + "text" + ], + "supports_audio_output": false, + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "supports_url_context": true, + "supports_vision": true, + "supports_web_search": true, + "tpm": 8000000, + "search_context_cost_per_query": { + "search_context_size_low": 0.035, + "search_context_size_medium": 0.035, + "search_context_size_high": 0.035 + } + }, + "vertex_ai/claude-sonnet-4-6@default": { + "cache_creation_input_token_cost": 3.75e-06, + "cache_read_input_token_cost": 3e-07, + "input_cost_per_token": 3e-06, + "litellm_provider": "vertex_ai-anthropic_models", + "max_input_tokens": 1000000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 1.5e-05, + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_pdf_input": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "search_context_cost_per_query": { + "search_context_size_high": 0.01, + "search_context_size_low": 0.01, + "search_context_size_medium": 0.01 + }, + "supports_minimal_reasoning_effort": true + }, + "duckduckgo/search": { + "litellm_provider": "duckduckgo", + "mode": "search", + "input_cost_per_query": 0.0, + "metadata": { + "notes": "DuckDuckGo Instant Answer API is free and does not require an API key." + } + }, + "bedrock_mantle/openai.gpt-oss-120b": { + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "bedrock_mantle", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "bedrock_mantle/openai.gpt-oss-20b": { + "input_cost_per_token": 7.5e-08, + "output_cost_per_token": 3e-07, + "litellm_provider": "bedrock_mantle", + "max_input_tokens": 131072, + "max_output_tokens": 32768, + "max_tokens": 32768, + "mode": "chat", + "supports_function_calling": true, + "supports_parallel_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "bedrock_mantle/openai.gpt-oss-safeguard-120b": { + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 6e-07, + "litellm_provider": "bedrock_mantle", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "bedrock_mantle/openai.gpt-oss-safeguard-20b": { + "input_cost_per_token": 7.5e-08, + "output_cost_per_token": 3e-07, + "litellm_provider": "bedrock_mantle", + "max_input_tokens": 131072, + "max_output_tokens": 65536, + "max_tokens": 65536, + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true + }, + "volcengine/doubao-seed-2-0-pro-260215": { + "litellm_provider": "volcengine", + "max_input_tokens": 256000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "source": "https://www.volcengine.com/docs/82379/1330310", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": false, + "supports_vision": true, + "tiered_pricing": [ + { + "input_cost_per_token": 4.6e-07, + "output_cost_per_token": 2.3e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 7e-07, + "output_cost_per_token": 3.5e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 1.4e-06, + "output_cost_per_token": 7e-06, + "range": [ + 128000.0, + 256000.0 + ] + } + ] + }, + "volcengine/doubao-seed-2-0-lite-260215": { + "litellm_provider": "volcengine", + "max_input_tokens": 256000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "source": "https://www.volcengine.com/docs/82379/1330310", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": false, + "supports_vision": true, + "tiered_pricing": [ + { + "input_cost_per_token": 8.7e-08, + "output_cost_per_token": 5.2e-07, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 1.3e-07, + "output_cost_per_token": 7.8e-07, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 2.6e-07, + "output_cost_per_token": 1.6e-06, + "range": [ + 128000.0, + 256000.0 + ] + } + ] + }, + "volcengine/doubao-seed-2-0-mini-260215": { + "litellm_provider": "volcengine", + "max_input_tokens": 256000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "source": "https://www.volcengine.com/docs/82379/1330310", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": false, + "supports_vision": true, + "tiered_pricing": [ + { + "input_cost_per_token": 2.9e-08, + "output_cost_per_token": 2.9e-07, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 5.8e-08, + "output_cost_per_token": 5.8e-07, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 1.2e-07, + "output_cost_per_token": 1.2e-06, + "range": [ + 128000.0, + 256000.0 + ] + } + ] + }, + "volcengine/doubao-seed-2-0-code-preview-260215": { + "litellm_provider": "volcengine", + "max_input_tokens": 256000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "source": "https://www.volcengine.com/docs/82379/1330310", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_tool_choice": false, + "supports_vision": true, + "tiered_pricing": [ + { + "input_cost_per_token": 4.6e-07, + "output_cost_per_token": 2.3e-06, + "range": [ + 0, + 32000.0 + ] + }, + { + "input_cost_per_token": 7e-07, + "output_cost_per_token": 3.5e-06, + "range": [ + 32000.0, + 128000.0 + ] + }, + { + "input_cost_per_token": 1.4e-06, + "output_cost_per_token": 7e-06, + "range": [ + 128000.0, + 256000.0 + ] + } + ] + }, + "zai.glm-5": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3.2e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-1/zai.glm-5": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3.2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-west-2/zai.glm-5": { + "input_cost_per_token": 1e-06, + "output_cost_per_token": 3.2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 128000, + "max_tokens": 128000, + "mode": "chat", + "supports_function_calling": true, + "supports_reasoning": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "bedrock_converse", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-east-1/minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-west-2/minimax.minimax-m2.5": { + "input_cost_per_token": 3e-07, + "output_cost_per_token": 1.2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 1000000, + "max_output_tokens": 8192, + "max_tokens": 8192, + "mode": "chat", + "supports_function_calling": true, + "supports_system_messages": true, + "supports_tool_choice": true, + "source": "https://aws.amazon.com/bedrock/pricing/" + }, + "bedrock/us-gov-east-1/anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.5e-06, + "cache_read_input_token_cost": 1.2e-07, + "input_cost_per_token": 1.2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_pdf_input": true + }, + "bedrock/us-gov-west-1/anthropic.claude-haiku-4-5-20251001-v1:0": { + "cache_creation_input_token_cost": 1.5e-06, + "cache_read_input_token_cost": 1.2e-07, + "input_cost_per_token": 1.2e-06, + "litellm_provider": "bedrock", + "max_input_tokens": 200000, + "max_output_tokens": 64000, + "max_tokens": 64000, + "mode": "chat", + "output_cost_per_token": 6e-06, + "source": "https://aws.amazon.com/about-aws/whats-new/2025/10/claude-4-5-haiku-anthropic-amazon-bedrock", + "supports_assistant_prefill": true, + "supports_computer_use": true, + "supports_function_calling": true, + "supports_prompt_caching": true, + "supports_reasoning": true, + "supports_response_schema": true, + "supports_tool_choice": true, + "supports_vision": true, + "tool_use_system_prompt_tokens": 346, + "supports_native_structured_output": true, + "supports_pdf_input": true + } +} diff --git a/crates/headroom-proxy/src/compression/anthropic.rs b/crates/headroom-proxy/src/compression/anthropic.rs new file mode 100644 index 000000000..3131f6097 --- /dev/null +++ b/crates/headroom-proxy/src/compression/anthropic.rs @@ -0,0 +1,325 @@ +//! Anthropic `/v1/messages` request compression. +//! +//! # Request shape (relevant subset) +//! +//! ```json +//! { +//! "model": "claude-3-5-sonnet-20241022", +//! "system": "...", // string OR list of blocks (optional) +//! "messages": [ +//! {"role": "user", "content": "..."}, +//! {"role": "assistant", "content": [...]} +//! ], +//! "tools": [...], // optional +//! "max_tokens": 1024, // required +//! ... +//! } +//! ``` +//! +//! # What we do +//! +//! 1. Parse the body as JSON. On failure → passthrough. +//! 2. Pull `messages` (the only field we touch). On absence → passthrough. +//! 3. Pull `model` and `max_tokens` to compute the available budget. +//! 4. Run the ICM's `should_apply` gate. Under-budget → passthrough. +//! 5. Run `apply()`. Re-insert the (possibly trimmed) `messages` into +//! the parsed JSON. Re-serialize. +//! 6. On *any* error along the way: passthrough with a warn log. The +//! proxy must never break a request because compression failed. +//! +//! # What we DON'T do +//! +//! - Touch `system`. Anthropic separates system from messages; our +//! ICM operates on the messages list. The system tokens are +//! "invisible" to ICM's budget calculation, which means we +//! under-count slightly — that's fine (we'll compress less than +//! strictly necessary, never more). +//! - Touch `tools`, `temperature`, `top_p`, etc. These pass through +//! verbatim because they're tiny and load-bearing for behaviour. +//! - Compress individual content blocks. That's content-router / +//! pipeline work, scoped to a follow-up PR. ICM operates at the +//! message-list level only. + +use bytes::Bytes; +use serde_json::Value; + +use headroom_core::context::{ApplyCtx, IntelligentContextManager}; + +use super::model_limits::context_window_for; + +/// What happened. Used for the request-level tracing log. +#[derive(Debug)] +pub enum Outcome { + /// Body was unchanged. Reasons listed in `reason`. + Passthrough { reason: PassthroughReason }, + /// ICM ran but didn't drop anything (already under budget). + NoCompression { tokens_before: usize }, + /// ICM ran and trimmed the message list. + Compressed { + body: Bytes, + tokens_before: usize, + tokens_after: usize, + strategies_applied: Vec<&'static str>, + markers_inserted: Vec, + }, +} + +/// Why we passed the body through unchanged. +#[derive(Debug, Clone, Copy)] +pub enum PassthroughReason { + /// JSON parse failed. + NotJson, + /// `messages` was missing or not a JSON array. + NoMessages, + /// Re-serialization of the modified body failed (shouldn't + /// happen — we just deserialized this shape). + SerializeFailed, +} + +/// Run ICM over an Anthropic-shape body. Returns one of: +/// +/// - `Outcome::Compressed` — caller should forward `outcome.body` +/// instead of the original bytes. +/// - `Outcome::NoCompression` — caller forwards the original +/// bytes; ICM's `should_apply` returned false. +/// - `Outcome::Passthrough` — same as `NoCompression` from the +/// caller's perspective, but the reason is parse/serialize-related. +/// +/// Never returns an error. Compression failures degrade to +/// passthrough; this is the proxy's safety contract. +pub fn maybe_compress(body: &Bytes, icm: &IntelligentContextManager) -> Outcome { + let mut parsed: Value = match serde_json::from_slice(body) { + Ok(v) => v, + Err(_) => { + return Outcome::Passthrough { + reason: PassthroughReason::NotJson, + }; + } + }; + + // Move the messages array out of the object so we can hand + // ownership to ICM. We re-insert at the end. If `messages` is + // missing or not an array, passthrough. + let messages = match parsed.get_mut("messages") { + Some(Value::Array(_)) => match parsed["messages"].take() { + Value::Array(a) => a, + _ => unreachable!("just matched as_array"), + }, + _ => { + return Outcome::Passthrough { + reason: PassthroughReason::NoMessages, + }; + } + }; + + let model = parsed + .get("model") + .and_then(Value::as_str) + .unwrap_or_default(); + // `context_window_for` returns `u32` (LiteLLM-sourced). ICM's + // `ApplyCtx::model_limit` wants `usize`. The cast is lossless on + // every platform we run on — context windows are far below 4GB. + let model_limit = context_window_for(model) as usize; + + // Anthropic requires `max_tokens`; if absent (malformed), assume + // a small reservation rather than zero so we don't pretend the + // whole window is available for input. + let output_buffer = parsed + .get("max_tokens") + .and_then(Value::as_u64) + .map(|v| v as usize) + .unwrap_or(4_096); + + // Cheap pre-check before the real apply call. Saves the cost of + // a full message-list traversal when the request is small. + if !icm.should_apply(&messages, model_limit, output_buffer) { + // Restore messages and return without compression. + parsed["messages"] = Value::Array(messages); + // Compute tokens_before from the re-inserted form for log + // accuracy. Cheap: it's the same walk should_apply already + // did, but we don't have the count back from that call. We + // skip the recount and return 0; the caller's log just + // shows "no_compression" without a number, which is fine. + return Outcome::NoCompression { tokens_before: 0 }; + } + + let result = icm.apply( + messages, + ApplyCtx { + model_limit, + output_buffer: Some(output_buffer), + // TODO: detect provider prefix-cached messages from the + // request. Anthropic exposes prompt caching via + // `cache_control` on content blocks. Until we wire that + // detection, we treat the whole list as droppable. + frozen_message_count: 0, + }, + ); + + // ICM may return tokens_after >= tokens_before when no drops + // happened (e.g. everything is protected). Treat that as + // no-compression rather than ship a needless re-serialize. + if result.tokens_after >= result.tokens_before { + // Reinsert the (unchanged) messages and report. + parsed["messages"] = Value::Array(result.messages); + return Outcome::NoCompression { + tokens_before: result.tokens_before, + }; + } + + parsed["messages"] = Value::Array(result.messages); + + let new_body = match serde_json::to_vec(&parsed) { + Ok(v) => Bytes::from(v), + Err(_) => { + return Outcome::Passthrough { + reason: PassthroughReason::SerializeFailed, + }; + } + }; + + Outcome::Compressed { + body: new_body, + tokens_before: result.tokens_before, + tokens_after: result.tokens_after, + strategies_applied: result.strategies_applied, + markers_inserted: result.markers_inserted, + } +} + +#[cfg(test)] +mod tests { + use super::*; + use crate::compression::icm::build_icm; + use serde_json::json; + + fn icm() -> std::sync::Arc { + build_icm().expect("ICM builds") + } + + #[test] + fn passthrough_on_invalid_json() { + let icm = icm(); + let body = Bytes::from_static(b"not json"); + match maybe_compress(&body, &icm) { + Outcome::Passthrough { + reason: PassthroughReason::NotJson, + } => {} + other => panic!("expected NotJson passthrough, got {other:?}"), + } + } + + #[test] + fn passthrough_when_messages_field_missing() { + let icm = icm(); + let body = Bytes::from(json!({"model": "claude-3-5-sonnet-20241022"}).to_string()); + match maybe_compress(&body, &icm) { + Outcome::Passthrough { + reason: PassthroughReason::NoMessages, + } => {} + other => panic!("expected NoMessages passthrough, got {other:?}"), + } + } + + #[test] + fn passthrough_when_messages_not_array() { + let icm = icm(); + let body = Bytes::from( + json!({"model": "claude-3-5-sonnet", "messages": "not-an-array"}).to_string(), + ); + match maybe_compress(&body, &icm) { + Outcome::Passthrough { + reason: PassthroughReason::NoMessages, + } => {} + other => panic!("expected NoMessages passthrough, got {other:?}"), + } + } + + #[test] + fn no_compression_when_under_budget() { + let icm = icm(); + let body = Bytes::from( + json!({ + "model": "claude-3-5-sonnet-20241022", + "max_tokens": 1024, + "messages": [{"role": "user", "content": "hello"}] + }) + .to_string(), + ); + match maybe_compress(&body, &icm) { + Outcome::NoCompression { .. } => {} + other => panic!("expected NoCompression, got {other:?}"), + } + } + + #[test] + fn compresses_when_over_budget() { + let icm = icm(); + // Squeeze the available budget by setting a huge max_tokens + // so output_buffer eats almost the whole window. With a + // 200K window for Claude and max_tokens=199_500, only ~500 + // tokens are available — anything bigger forces compression. + let big_messages: Vec = (0..30) + .map(|i| { + json!({ + "role": if i % 2 == 0 { "user" } else { "assistant" }, + "content": format!("padding token {i} ").repeat(20), + }) + }) + .collect(); + let body = Bytes::from( + json!({ + "model": "claude-3-5-sonnet-20241022", + "max_tokens": 199_500, + "messages": big_messages, + }) + .to_string(), + ); + match maybe_compress(&body, &icm) { + Outcome::Compressed { + body: new_body, + tokens_before, + tokens_after, + .. + } => { + assert!(tokens_after < tokens_before); + // The new body is valid JSON with a shorter messages + // array (or includes the CCR marker that the shim's + // injection logic adds — but the proxy doesn't do + // marker injection; that lives in the Python shim). + let parsed: Value = serde_json::from_slice(&new_body).unwrap(); + assert!(parsed["messages"].as_array().is_some()); + } + other => panic!("expected Compressed, got {other:?}"), + } + } + + #[test] + fn unknown_model_does_not_panic() { + // Should fall back to the default 128K window and behave + // like any other request. + let icm = icm(); + let body = Bytes::from( + json!({ + "model": "future-model-2099", + "max_tokens": 100, + "messages": [{"role": "user", "content": "hi"}] + }) + .to_string(), + ); + let _ = maybe_compress(&body, &icm); // shouldn't panic + } + + #[test] + fn missing_max_tokens_does_not_panic() { + let icm = icm(); + let body = Bytes::from( + json!({ + "model": "claude-3-5-sonnet-20241022", + "messages": [{"role": "user", "content": "hi"}] + }) + .to_string(), + ); + let _ = maybe_compress(&body, &icm); // shouldn't panic + } +} diff --git a/crates/headroom-proxy/src/compression/icm.rs b/crates/headroom-proxy/src/compression/icm.rs new file mode 100644 index 000000000..2c34fcb25 --- /dev/null +++ b/crates/headroom-proxy/src/compression/icm.rs @@ -0,0 +1,70 @@ +//! Build the per-process `IntelligentContextManager`. +//! +//! Constructed once at proxy startup, stored in `AppState`, shared +//! across every request via `Arc`. The `MessageScorer` inside has a +//! `Mutex` embedding cache; contention is low because we +//! aren't wiring an `EmbeddingProvider` yet (that's a follow-up). +//! +//! # Tokenizer choice +//! +//! `IntelligentContextManager` needs a `Tokenizer` to count tokens +//! for the `should_apply` budget gate. We use `TiktokenCounter` with +//! the `gpt-4o-mini` (`o200k_base`) encoding because: +//! +//! - It's the most modern tiktoken vocabulary that's broadly +//! compatible with both OpenAI and Anthropic message shapes. +//! - It's strictly more accurate than the `EstimatingCounter` +//! (chars/4) default that the PyO3 binding hardcodes. +//! - For Anthropic specifically, the actual tokenizer is bespoke and +//! not publicly available; tiktoken-based counting is the +//! industry-standard close-enough estimate. Errors are O(±5%) +//! which doesn't change `should_apply` decisions. +//! +//! # CCR store +//! +//! An `InMemoryCcrStore` is constructed alongside the manager. When +//! ICM drops messages, their original JSON gets stashed under a +//! content-hash key and a marker is inserted into the surviving +//! message stream. If the LLM later calls a `ccr_retrieve` tool, the +//! handler can serve the dropped content from this store. (The +//! retrieval-handling tool is a separate concern; we just construct +//! the store so drops are recoverable.) + +use std::sync::Arc; + +use headroom_core::ccr::{CcrStore, InMemoryCcrStore}; +use headroom_core::context::{IcmConfig, IntelligentContextManager}; +use headroom_core::tokenizer::{TiktokenCounter, Tokenizer}; + +/// Construct the proxy's shared ICM. Returns `Arc` for cheap cloning +/// into request handlers. +/// +/// Errors only on tokenizer construction — `gpt-4o-mini` is always +/// available since tiktoken-rs ships its vocabulary, but we bubble +/// up the error path for symmetry with the rest of the proxy's +/// fallible startup. +pub fn build_icm() -> Result, String> { + let tokenizer: Arc = Arc::new( + TiktokenCounter::for_model("gpt-4o-mini") + .map_err(|e| format!("init TiktokenCounter for gpt-4o-mini: {e}"))?, + ); + let ccr: Arc = Arc::new(InMemoryCcrStore::new()); + Ok(Arc::new(IntelligentContextManager::new( + IcmConfig::default(), + tokenizer, + Some(ccr), + ))) +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn build_icm_succeeds() { + let icm = build_icm().expect("ICM should build"); + // Smoke: should_apply on an empty list under a generous + // budget returns false (no work). + assert!(!icm.should_apply(&[], 128_000, 4_000)); + } +} diff --git a/crates/headroom-proxy/src/compression/mod.rs b/crates/headroom-proxy/src/compression/mod.rs new file mode 100644 index 000000000..f80979dc9 --- /dev/null +++ b/crates/headroom-proxy/src/compression/mod.rs @@ -0,0 +1,70 @@ +//! Compression interceptor for LLM-shaped requests. +//! +//! The proxy is a streaming reverse proxy by default. When +//! `--compression` is enabled and a request hits a known LLM +//! provider path, we buffer the body, run +//! `IntelligentContextManager` over the message list, and forward +//! the (possibly trimmed) body upstream. Everything else stays +//! streaming, including: +//! +//! - WebSocket upgrades — handled in the catch-all before +//! `forward_http` is called; never reach this module. +//! - Non-LLM paths (any URL not matching a known provider). +//! - Non-JSON content types (skip; we don't speculate at body +//! contents we don't know how to parse). +//! - Streaming SSE responses — only the request body is touched; +//! responses pass through untouched. +//! +//! # Provider matrix (current + planned) +//! +//! | Provider | Path | Status | +//! |--------------|-----------------------|--------| +//! | Anthropic | `POST /v1/messages` | ✅ this module | +//! | OpenAI | `POST /v1/chat/completions` | follow-up | +//! | Google | `POST /v1beta/...` | follow-up | +//! | Bedrock | varied | follow-up | +//! +//! # Failure-mode contract +//! +//! Compression must NEVER break a request. Every error path — +//! parse failure, missing field, body too large, unknown model — +//! falls through to the original body being forwarded unchanged. +//! Operators see what happened in `tracing` warnings; clients see +//! their request go through. + +pub mod anthropic; +pub mod icm; +pub mod model_limits; + +pub use anthropic::{maybe_compress, Outcome, PassthroughReason}; +pub use icm::build_icm; + +/// Does this request path target an LLM endpoint we know how to +/// compress? Cheap pre-filter before buffering the body. +pub fn is_compressible_path(path: &str) -> bool { + // Exact-match the Anthropic Messages endpoint. Future providers + // get their own arms here. Avoid prefix-matching to keep the + // compression scope explicit — `/v1/messages/123` (a + // hypothetical future per-message endpoint) shouldn't accidentally + // get its body parsed as a chat-completions request. + path == "/v1/messages" +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn anthropic_messages_path_matches() { + assert!(is_compressible_path("/v1/messages")); + } + + #[test] + fn other_paths_skip() { + assert!(!is_compressible_path("/v1/messages/123")); + assert!(!is_compressible_path("/v1/chat/completions")); + assert!(!is_compressible_path("/healthz")); + assert!(!is_compressible_path("/")); + assert!(!is_compressible_path("")); + } +} diff --git a/crates/headroom-proxy/src/compression/model_limits.rs b/crates/headroom-proxy/src/compression/model_limits.rs new file mode 100644 index 000000000..e4e64471c --- /dev/null +++ b/crates/headroom-proxy/src/compression/model_limits.rs @@ -0,0 +1,211 @@ +//! Model name → context window (tokens) lookup, sourced from LiteLLM. +//! +//! # Why LiteLLM and not a hand-rolled table +//! +//! Earlier drafts of this module hardcoded a small `if/else` chain +//! covering Claude / GPT-4o / GPT-3.5. Two failure modes that cost us: +//! +//! 1. **Static rot.** Anthropic and OpenAI ship new models monthly. +//! A hardcoded table goes stale between Headroom releases; new +//! models silently fall through to a default that is often wrong. +//! 2. **Long tail.** Bedrock, Cohere, Mistral, Gemini, AzureML — each +//! has dozens of model variants. We aren't going to maintain +//! a comprehensive table by hand. +//! +//! [LiteLLM] maintains `model_prices_and_context_window.json` as a +//! community-curated source of truth: ~2000 chat models, refreshed +//! weekly, with `max_input_tokens` (the field we care about) plus +//! pricing, output limits, capability flags. Every other LLM-tool +//! ecosystem (Portkey, Helicone, OpenRouter, Continue) pulls from it. +//! +//! [LiteLLM]: https://github.com/BerriAI/litellm +//! +//! # Vendoring strategy +//! +//! We **check the JSON into the repo** at +//! `crates/headroom-proxy/data/model_prices_and_context_window.json` +//! and `include_str!` it at compile time. Reasons over alternatives: +//! +//! - **Build-time fetch (`build.rs` + curl):** breaks for offline / +//! air-gapped builds — bad for BYOC where customers may not allow +//! outbound network during install. +//! - **Runtime fetch:** same problem, plus a startup-failure surface +//! we don't need. +//! +//! Refresh is operator-driven: `scripts/refresh_model_limits.sh` +//! re-pulls and validates the JSON, the diff lands in a regular PR. +//! We trade "always fresh" for "deterministic, offline-buildable, +//! auditable in version control" — the right trade for a deploy +//! artifact that's expected to run in customer VPCs. +//! +//! # Performance +//! +//! The 1.4MB JSON parses in ~10ms one-time on first lookup. Parsed +//! result is cached in a `OnceLock` so subsequent lookups +//! are O(1). When `--compression` is off, the JSON is in the binary +//! image but never parsed — zero runtime cost. +//! +//! # Default for unknown models +//! +//! When a model isn't in the table, we return a conservative 128K +//! and emit a `tracing::warn!` (once per unknown model id). 128K is +//! the dominant context window across modern frontier models; being +//! wrong here means we either over-compress (safe — we just trim +//! unnecessarily, the request still works) or under-compress (the +//! upstream rejects it with `context_length_exceeded`, which is +//! recoverable by the client). + +use std::collections::HashMap; +use std::sync::OnceLock; + +/// Conservative default for unknown models. Modern frontier models +/// almost universally have ≥128K context; we err on the side of +/// over-compressing (safe) rather than under-compressing (broken). +pub(crate) const DEFAULT_CONTEXT_WINDOW: u32 = 128_000; + +/// LiteLLM's vendored model price + context-window table. Refreshed +/// via `scripts/refresh_model_limits.sh`. ~1.4MB; embedded into the +/// binary so the proxy ships with no startup network dependency. +const VENDORED_JSON: &str = include_str!("../../data/model_prices_and_context_window.json"); + +/// Parsed lookup: model id → max input tokens. Built lazily on +/// first call; subsequent calls reuse the same `HashMap`. +static TABLE: OnceLock> = OnceLock::new(); + +/// Looks up `max_input_tokens` for `model`. Returns +/// [`DEFAULT_CONTEXT_WINDOW`] when the model isn't in the table. +/// +/// Lookup is exact-match by model id. We deliberately do NOT do +/// prefix matching — model versions are semantically distinct +/// (`claude-3-5-sonnet-20241022` was 200K, but a hypothetical +/// `claude-3-5-sonnet-mini` may be different) and a prefix rule +/// would cause silent wrong answers. +pub fn context_window_for(model: &str) -> u32 { + let table = TABLE.get_or_init(parse_vendored); + if let Some(&n) = table.get(model) { + return n; + } + // Unknown model. We don't log here on every miss — that's + // per-request noise. The caller (compression::anthropic) logs + // once, with the model id, when this happens. Just return the + // default and let the caller handle observability. + DEFAULT_CONTEXT_WINDOW +} + +/// Walk the LiteLLM JSON and extract the chat-model context windows. +/// +/// LiteLLM's schema: top-level object whose keys are model ids. +/// Values may be: +/// - `sample_spec` — a template entry; skipped. +/// - Image / audio / embedding entries — `mode != "chat"`; skipped. +/// - Chat entries — have `max_input_tokens` (preferred) or +/// `max_tokens` (legacy fallback). +/// +/// The 79-ish chat entries in the current snapshot that lack BOTH +/// fields are skipped silently — they'd hit `DEFAULT_CONTEXT_WINDOW` +/// at lookup time anyway. +fn parse_vendored() -> HashMap { + let raw: serde_json::Value = serde_json::from_str(VENDORED_JSON) + .expect("vendored LiteLLM JSON must parse at build time"); + let obj = raw + .as_object() + .expect("LiteLLM JSON must be a top-level object"); + + // Slight over-allocation; better than reallocating during the walk. + let mut out: HashMap = HashMap::with_capacity(obj.len()); + for (key, val) in obj { + if key == "sample_spec" { + continue; + } + let entry = match val.as_object() { + Some(o) => o, + None => continue, + }; + // Only chat-mode models are relevant for our compressor. + // Image / audio / embedding endpoints don't have a "messages" + // array we can compress. + if entry.get("mode").and_then(|m| m.as_str()) != Some("chat") { + continue; + } + + // Prefer max_input_tokens. Fall back to max_tokens (older + // entries used max_tokens as a synonym for input window). + let n = entry + .get("max_input_tokens") + .and_then(|v| v.as_u64()) + .or_else(|| entry.get("max_tokens").and_then(|v| v.as_u64())); + let Some(n) = n else { continue }; + + // u32 fits every realistic context window. The largest known + // today is ~10M (Magic.dev, hypothetical) — still under + // 4 billion. If a future model crosses u32::MAX we have + // larger problems than this `as`. + out.insert(key.clone(), n.min(u32::MAX as u64) as u32); + } + out +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn vendored_json_parses_at_runtime() { + // Calling this once forces parse via OnceLock. If the JSON + // is malformed, this test panics with a useful error before + // any lookup test runs. Subsequent tests in this module + // share the same parsed table. + let table = TABLE.get_or_init(parse_vendored); + assert!( + table.len() > 100, + "expected >100 chat models in LiteLLM snapshot, got {}", + table.len() + ); + } + + #[test] + fn current_claude_models_present() { + // Lock against the snapshot rotting silently. If LiteLLM + // renames the canonical Claude entry we want a test failure + // — not a silent fall-through to DEFAULT_CONTEXT_WINDOW. + // Pick a model we expect to remain stable: claude-sonnet-4-5 + // (current as of the snapshot fetch). + let n = context_window_for("claude-sonnet-4-5-20250929"); + assert_eq!(n, 200_000, "claude-sonnet-4-5 should be 200K input window"); + } + + #[test] + fn current_gpt_models_present() { + assert_eq!(context_window_for("gpt-4o-mini"), 128_000); + assert_eq!(context_window_for("gpt-4-turbo"), 128_000); + } + + #[test] + fn unknown_model_returns_default() { + assert_eq!( + context_window_for("definitely-not-a-real-model-2099"), + DEFAULT_CONTEXT_WINDOW + ); + assert_eq!(context_window_for(""), DEFAULT_CONTEXT_WINDOW); + } + + #[test] + fn empty_or_garbage_string_does_not_panic() { + // The lookup must not panic on adversarial input — bad + // model strings come from the wire and we forward unknown + // ones rather than failing the request. + let _ = context_window_for(""); + let _ = context_window_for("\0\0\0"); + let _ = context_window_for(&"x".repeat(10_000)); + } + + #[test] + fn sample_spec_entry_is_excluded() { + // LiteLLM's JSON includes a "sample_spec" template entry + // documenting the schema. It must not appear as a real + // model in our lookup — a request specifying it would + // otherwise get a bogus context window. + let table = TABLE.get_or_init(parse_vendored); + assert!(!table.contains_key("sample_spec")); + } +} diff --git a/crates/headroom-proxy/src/config.rs b/crates/headroom-proxy/src/config.rs index b0fe19032..6f03e0a6a 100644 --- a/crates/headroom-proxy/src/config.rs +++ b/crates/headroom-proxy/src/config.rs @@ -50,6 +50,28 @@ pub struct CliArgs { /// Maximum time to wait for in-flight requests to finish on shutdown. #[arg(long, default_value = "30s", value_parser = parse_duration)] pub graceful_shutdown_timeout: Duration, + + /// Enable Headroom compression on LLM-shaped requests + /// (currently: `POST /v1/messages` for Anthropic). When off, + /// the proxy stays a pure streaming passthrough. + /// + /// Off by default so existing operators get unchanged behaviour + /// and the integration-test harness doesn't need to opt out + /// per-test. Operators wanting to demo the compressor pass + /// `--compression` (or set `HEADROOM_PROXY_COMPRESSION=1`). + #[arg( + long = "compression", + env = "HEADROOM_PROXY_COMPRESSION", + default_value_t = false + )] + pub compression: bool, + + /// Maximum body size to buffer for compression. Bodies larger + /// than this get forwarded unchanged. Defaults to `--max-body-bytes` + /// when unset, so operators only need to tune one knob unless + /// they have a specific reason to cap compression separately. + #[arg(long, value_parser = parse_bytes)] + pub compression_max_body_bytes: Option, } fn parse_duration(s: &str) -> Result { @@ -73,6 +95,13 @@ pub struct Config { pub log_level: String, pub rewrite_host: bool, pub graceful_shutdown_timeout: Duration, + /// Master switch for the LLM compression interceptor. When `false`, + /// the proxy is pure streaming passthrough and never buffers a body. + pub compression: bool, + /// Effective ceiling for compression-time body buffering. + /// Inherits `max_body_bytes` when not overridden. Bodies larger + /// than this still forward, just unchanged. + pub compression_max_body_bytes: u64, } impl Config { @@ -82,6 +111,9 @@ impl Config { } else { args.rewrite_host }; + let compression_max_body_bytes = args + .compression_max_body_bytes + .unwrap_or(args.max_body_bytes); Self { listen: args.listen, upstream: args.upstream, @@ -91,10 +123,13 @@ impl Config { log_level: args.log_level, rewrite_host, graceful_shutdown_timeout: args.graceful_shutdown_timeout, + compression: args.compression, + compression_max_body_bytes, } } - /// Test/library helper. + /// Test/library helper. Compression off by default — match + /// production-default behaviour so existing tests stay unchanged. pub fn for_test(upstream: Url) -> Self { Self { listen: "127.0.0.1:0".parse().unwrap(), @@ -105,6 +140,8 @@ impl Config { log_level: "warn".into(), rewrite_host: true, graceful_shutdown_timeout: Duration::from_secs(5), + compression: false, + compression_max_body_bytes: 100 * 1024 * 1024, } } } diff --git a/crates/headroom-proxy/src/error.rs b/crates/headroom-proxy/src/error.rs index 25add9a03..b7a09729f 100644 --- a/crates/headroom-proxy/src/error.rs +++ b/crates/headroom-proxy/src/error.rs @@ -20,6 +20,16 @@ pub enum ProxyError { #[error("io error: {0}")] Io(#[from] std::io::Error), + + /// Surfaced when `--compression` is enabled but the proxy can't + /// build the IntelligentContextManager at startup (e.g. the + /// embedded tokenizer asset failed to initialize). Bubbles up to + /// `main` as a fatal startup error rather than a per-request + /// failure — if compression is configured but the engine won't + /// build, the operator should know immediately, not at first + /// LLM request. + #[error("compression engine startup failed: {0}")] + CompressionStartup(String), } impl IntoResponse for ProxyError { @@ -38,6 +48,12 @@ impl IntoResponse for ProxyError { ProxyError::InvalidHeader(_) => (StatusCode::BAD_REQUEST, self.to_string()), ProxyError::WebSocket(_) => (StatusCode::BAD_GATEWAY, self.to_string()), ProxyError::Io(_) => (StatusCode::INTERNAL_SERVER_ERROR, self.to_string()), + // CompressionStartup is a startup-time error, not a + // per-request one — but if it ever surfaces in the + // handler path, surface as 500 rather than panic. + ProxyError::CompressionStartup(_) => { + (StatusCode::INTERNAL_SERVER_ERROR, self.to_string()) + } }; tracing::warn!(error = %msg, "proxy error"); (status, msg).into_response() diff --git a/crates/headroom-proxy/src/lib.rs b/crates/headroom-proxy/src/lib.rs index cc9c347a0..3179c9fdc 100644 --- a/crates/headroom-proxy/src/lib.rs +++ b/crates/headroom-proxy/src/lib.rs @@ -1,6 +1,7 @@ //! headroom-proxy library: transparent reverse proxy in front of the Python //! Headroom proxy. Used by both `main.rs` and the integration tests. +pub mod compression; pub mod config; pub mod error; pub mod headers; diff --git a/crates/headroom-proxy/src/proxy.rs b/crates/headroom-proxy/src/proxy.rs index 91bf158b9..66859e9fc 100644 --- a/crates/headroom-proxy/src/proxy.rs +++ b/crates/headroom-proxy/src/proxy.rs @@ -4,7 +4,7 @@ use std::net::SocketAddr; use std::sync::Arc; use std::time::Instant; -use axum::body::Body; +use axum::body::{to_bytes, Body}; use axum::extract::{ConnectInfo, State, WebSocketUpgrade}; use axum::http::{HeaderMap, HeaderName, Request, Response, StatusCode, Uri}; use axum::response::IntoResponse; @@ -16,6 +16,9 @@ use futures_util::{StreamExt as _, TryStreamExt}; #[cfg(test)] use http_body_util::BodyExt; +use headroom_core::context::IntelligentContextManager; + +use crate::compression; use crate::config::Config; use crate::error::ProxyError; use crate::headers::{build_forward_request_headers, filter_response_headers}; @@ -27,6 +30,11 @@ use crate::websocket::ws_handler; pub struct AppState { pub config: Arc, pub client: reqwest::Client, + /// Optional shared `IntelligentContextManager`. Constructed only + /// when `config.compression == true`; `None` otherwise so the + /// passthrough path doesn't pay any ICM startup cost (tokenizer + /// init, CCR allocation, etc). + pub icm: Option>, } impl AppState { @@ -41,9 +49,22 @@ impl AppState { // Both HTTP/1.1 and HTTP/2 negotiated via ALPN. .build() .map_err(ProxyError::Upstream)?; + + // Construct ICM only when compression is enabled. ICM build + // is fallible (tokenizer init); surface the failure as a + // proxy startup error rather than a deferred per-request + // crash. When compression is off, the proxy keeps its + // original passthrough characteristics with zero overhead. + let icm = if config.compression { + Some(compression::build_icm().map_err(ProxyError::CompressionStartup)?) + } else { + None + }; + Ok(Self { config: Arc::new(config), client, + icm, }) } } @@ -81,6 +102,23 @@ async fn catch_all( .unwrap_or_else(|e| e.into_response()) } +/// True if `Content-Type` is `application/json` (with any optional +/// parameters like `; charset=utf-8`). Compression only inspects JSON +/// bodies — multipart uploads, form-encoded posts, and binary +/// payloads stream through untouched. +fn is_application_json(headers: &HeaderMap) -> bool { + headers + .get(http::header::CONTENT_TYPE) + .and_then(|v| v.to_str().ok()) + .map(|s| { + // Take the media-type portion before any ';'. Trim and + // compare case-insensitively per RFC 7231 §3.1.1.1. + let media_type = s.split(';').next().unwrap_or("").trim(); + media_type.eq_ignore_ascii_case("application/json") + }) + .unwrap_or(false) +} + fn is_websocket_upgrade(headers: &HeaderMap) -> bool { let upgrade = headers .get(http::header::UPGRADE) @@ -166,20 +204,133 @@ async fn forward_http( } } - // Stream the request body through to reqwest. We don't buffer. - let body_stream = - TryStreamExt::map_err(req.into_body().into_data_stream(), std::io::Error::other); - let reqwest_body = reqwest::Body::wrap_stream(body_stream); + // ─── COMPRESSION GATE ────────────────────────────────────────────── + // + // Streaming-by-default is the proxy's contract for everything that + // isn't explicitly an LLM-shape request. To inspect a body for + // compression we have to buffer, which is incompatible with + // streaming — so we make the buffering decision *here*, on a + // narrow gate: + // + // - Compression enabled in config? (`state.config.compression`) + // - Method is POST? (we only compress request bodies) + // - Path matches a known LLM endpoint? (`compression::is_compressible_path`) + // - Content-Type is application/json? (skip multipart, form, binary) + // - ICM was successfully built? (Some by construction when compression is on) + // + // ALL of those true → buffer + run ICM + forward modified body. + // ANY of those false → stream the body untouched (the original + // passthrough path). This keeps WebSocket upgrades, healthchecks, + // tool-API endpoints, and SSE streaming from paying any + // buffering cost. + let should_compress = state.config.compression + && method == axum::http::Method::POST + && compression::is_compressible_path(uri.path()) + && is_application_json(req.headers()) + && state.icm.is_some(); let reqwest_method = reqwest::Method::from_bytes(method.as_str().as_bytes()) .map_err(|e| ProxyError::InvalidHeader(e.to_string()))?; - let upstream_resp = state - .client - .request(reqwest_method, upstream_url.clone()) - .headers(outgoing_headers) - .body(reqwest_body) - .send() - .await?; + + let upstream_resp = if should_compress { + // Buffer up to `compression_max_body_bytes`. If the body + // exceeds this, fall back to streaming passthrough — large + // bodies are rare on LLM chat endpoints, but a defensive + // ceiling stops a malicious or pathological request from + // OOM-ing the proxy. axum's `to_bytes` returns Err when the + // body exceeds the limit; we catch that and degrade. + let max = state.config.compression_max_body_bytes as usize; + let buffered = match to_bytes(req.into_body(), max).await { + Ok(b) => b, + Err(e) => { + tracing::warn!( + request_id = %request_id, + path = %path_for_log, + limit_bytes = max, + error = %e, + "compression: body exceeds limit, falling back to streaming passthrough \ + is impossible (body already partially consumed) — failing the request", + ); + // Once `req.into_body()` is consumed by `to_bytes` we + // can no longer stream. The defensive choice is to + // fail the request loudly. Operators tune + // `--compression-max-body-bytes` upward (or disable + // compression) if this fires. + return Err(ProxyError::InvalidHeader(format!( + "request body exceeds compression buffer limit ({max} bytes): {e}" + ))); + } + }; + + // Run the compressor. Failures degrade to passthrough by + // returning the original buffered bytes. + let icm = state.icm.as_ref().expect("checked above"); + let outcome = compression::maybe_compress(&buffered, icm); + + let body_to_send = match outcome { + compression::Outcome::Compressed { + body, + tokens_before, + tokens_after, + strategies_applied, + markers_inserted, + } => { + tracing::info!( + request_id = %request_id, + path = %path_for_log, + tokens_before = tokens_before, + tokens_after = tokens_after, + tokens_freed = tokens_before.saturating_sub(tokens_after), + strategies = ?strategies_applied, + markers = markers_inserted.len(), + "compression applied" + ); + body + } + compression::Outcome::NoCompression { tokens_before } => { + tracing::debug!( + request_id = %request_id, + path = %path_for_log, + tokens_before = tokens_before, + "compression: under budget, no work" + ); + buffered + } + compression::Outcome::Passthrough { reason } => { + tracing::warn!( + request_id = %request_id, + path = %path_for_log, + reason = ?reason, + "compression: passthrough on parse/serialize" + ); + buffered + } + }; + + // Forward the (possibly modified) body. reqwest sets its own + // Content-Length from the body bytes — the existing + // `build_forward_request_headers` already strips the + // client-supplied Content-Length for us. + state + .client + .request(reqwest_method, upstream_url.clone()) + .headers(outgoing_headers) + .body(body_to_send) + .send() + .await? + } else { + // Pure streaming path — the original passthrough behaviour. + let body_stream = + TryStreamExt::map_err(req.into_body().into_data_stream(), std::io::Error::other); + let reqwest_body = reqwest::Body::wrap_stream(body_stream); + state + .client + .request(reqwest_method, upstream_url.clone()) + .headers(outgoing_headers) + .body(reqwest_body) + .send() + .await? + }; let upstream_status = upstream_resp.status(); let status = StatusCode::from_u16(upstream_status.as_u16()).unwrap_or(StatusCode::BAD_GATEWAY); diff --git a/crates/headroom-proxy/tests/common/mod.rs b/crates/headroom-proxy/tests/common/mod.rs index d4fa008dc..96b41c4b2 100644 --- a/crates/headroom-proxy/tests/common/mod.rs +++ b/crates/headroom-proxy/tests/common/mod.rs @@ -34,8 +34,20 @@ impl ProxyHandle { #[allow(dead_code)] pub async fn start_proxy(upstream: &str) -> ProxyHandle { + start_proxy_with(upstream, |_| {}).await +} + +/// Start a proxy with a customized `Config`. The closure receives a +/// mutable reference to the default `Config::for_test` and may toggle +/// flags like `compression` before the proxy is built. +#[allow(dead_code)] +pub async fn start_proxy_with(upstream: &str, customize: F) -> ProxyHandle +where + F: FnOnce(&mut Config), +{ let upstream_url: Url = upstream.parse().expect("valid upstream url"); - let config = Config::for_test(upstream_url); + let mut config = Config::for_test(upstream_url); + customize(&mut config); let state = AppState::new(config.clone()).expect("app state"); let app = build_app(state).into_make_service_with_connect_info::(); let listener = tokio::net::TcpListener::bind("127.0.0.1:0") diff --git a/crates/headroom-proxy/tests/integration_compression.rs b/crates/headroom-proxy/tests/integration_compression.rs new file mode 100644 index 000000000..850eeac51 --- /dev/null +++ b/crates/headroom-proxy/tests/integration_compression.rs @@ -0,0 +1,216 @@ +//! End-to-end integration tests for the compression interceptor. +//! +//! These tests boot a real Rust proxy in front of a wiremock upstream +//! and verify the request body that arrives at the upstream — i.e. we +//! observe the *actual* compression effect on the wire, not the +//! library outcome in isolation. +//! +//! Coverage: +//! - Compression-off (default): proxy is a passthrough, body is byte-identical. +//! - Compression-on, small body: ICM short-circuits, body unchanged. +//! - Compression-on, oversized body: ICM trims the messages array. +//! - Compression-on, non-JSON body: skipped (Content-Type gate). +//! - Compression-on, non-LLM path: skipped (path gate). + +mod common; + +use common::start_proxy_with; +use serde_json::{json, Value}; +use std::sync::{Arc, Mutex}; +use wiremock::matchers::{method, path}; +use wiremock::{Mock, MockServer, ResponseTemplate}; + +/// Mount a /v1/messages handler that captures the upstream request body +/// into the returned Arc> for assertions, and returns 200 OK. +async fn mount_anthropic_capture(upstream: &MockServer) -> Arc>>> { + let captured: Arc>>> = Arc::new(Mutex::new(None)); + let captured_clone = captured.clone(); + Mock::given(method("POST")) + .and(path("/v1/messages")) + .respond_with(move |req: &wiremock::Request| { + *captured_clone.lock().unwrap() = Some(req.body.clone()); + ResponseTemplate::new(200).set_body_string(r#"{"ok":true}"#) + }) + .mount(upstream) + .await; + captured +} + +/// Build a payload that's large enough to force ICM to trim. Uses the +/// same pattern as the `compresses_when_over_budget` unit test in the +/// anthropic module: huge max_tokens eats the budget, leaving very few +/// tokens for input and forcing drops. +fn oversized_anthropic_payload() -> Value { + let messages: Vec = (0..30) + .map(|i| { + json!({ + "role": if i % 2 == 0 { "user" } else { "assistant" }, + "content": format!("padding token {i} ").repeat(20), + }) + }) + .collect(); + json!({ + "model": "claude-3-5-sonnet-20241022", + "max_tokens": 199_500, + "messages": messages, + }) +} + +#[tokio::test] +async fn compression_off_passes_body_unchanged() { + let upstream = MockServer::start().await; + let captured = mount_anthropic_capture(&upstream).await; + let proxy = start_proxy_with(&upstream.uri(), |_| { + // compression remains off (Config::for_test default) + }) + .await; + + let payload = oversized_anthropic_payload(); + let body = serde_json::to_vec(&payload).unwrap(); + let resp = reqwest::Client::new() + .post(format!("{}/v1/messages", proxy.url())) + .header("content-type", "application/json") + .body(body.clone()) + .send() + .await + .unwrap(); + assert_eq!(resp.status(), 200); + + let got = captured.lock().unwrap().clone().expect("upstream got body"); + assert_eq!(got, body, "compression off — body must be byte-identical"); + proxy.shutdown().await; +} + +#[tokio::test] +async fn compression_on_short_body_passes_through() { + let upstream = MockServer::start().await; + let captured = mount_anthropic_capture(&upstream).await; + let proxy = start_proxy_with(&upstream.uri(), |c| { + c.compression = true; + }) + .await; + + let payload = json!({ + "model": "claude-3-5-sonnet-20241022", + "max_tokens": 1024, + "messages": [{"role": "user", "content": "hello"}], + }); + let body = serde_json::to_vec(&payload).unwrap(); + let resp = reqwest::Client::new() + .post(format!("{}/v1/messages", proxy.url())) + .header("content-type", "application/json") + .body(body.clone()) + .send() + .await + .unwrap(); + assert_eq!(resp.status(), 200); + + let got = captured.lock().unwrap().clone().expect("upstream got body"); + let got_json: Value = serde_json::from_slice(&got).unwrap(); + let in_messages = payload["messages"].as_array().unwrap().len(); + let out_messages = got_json["messages"].as_array().unwrap().len(); + assert_eq!( + out_messages, in_messages, + "small request stays under budget; messages array unchanged" + ); + proxy.shutdown().await; +} + +#[tokio::test] +async fn compression_on_oversized_body_trims_messages() { + let upstream = MockServer::start().await; + let captured = mount_anthropic_capture(&upstream).await; + let proxy = start_proxy_with(&upstream.uri(), |c| { + c.compression = true; + }) + .await; + + let payload = oversized_anthropic_payload(); + let body = serde_json::to_vec(&payload).unwrap(); + let in_messages = payload["messages"].as_array().unwrap().len(); + let resp = reqwest::Client::new() + .post(format!("{}/v1/messages", proxy.url())) + .header("content-type", "application/json") + .body(body.clone()) + .send() + .await + .unwrap(); + assert_eq!(resp.status(), 200); + + let got = captured.lock().unwrap().clone().expect("upstream got body"); + assert_ne!(got, body, "ICM should have trimmed something"); + let got_json: Value = serde_json::from_slice(&got).unwrap(); + let out_messages = got_json["messages"].as_array().unwrap().len(); + assert!( + out_messages < in_messages, + "expected fewer messages after compression: in={in_messages}, out={out_messages}" + ); + // Other fields preserved verbatim. + assert_eq!(got_json["model"], payload["model"]); + assert_eq!(got_json["max_tokens"], payload["max_tokens"]); + proxy.shutdown().await; +} + +#[tokio::test] +async fn compression_on_non_json_skips() { + let upstream = MockServer::start().await; + let captured = mount_anthropic_capture(&upstream).await; + let proxy = start_proxy_with(&upstream.uri(), |c| { + c.compression = true; + }) + .await; + + // Path matches /v1/messages but Content-Type isn't JSON. The gate + // must skip and stream verbatim — even though the body would + // otherwise be massive enough to compress. + let body = vec![0xAAu8; 64 * 1024]; + let resp = reqwest::Client::new() + .post(format!("{}/v1/messages", proxy.url())) + .header("content-type", "application/octet-stream") + .body(body.clone()) + .send() + .await + .unwrap(); + assert_eq!(resp.status(), 200); + + let got = captured.lock().unwrap().clone().expect("upstream got body"); + assert_eq!(got, body, "non-JSON content-type must bypass compression"); + proxy.shutdown().await; +} + +#[tokio::test] +async fn compression_on_non_llm_path_skips() { + let upstream = MockServer::start().await; + let captured: Arc>>> = Arc::new(Mutex::new(None)); + let captured_clone = captured.clone(); + Mock::given(method("POST")) + .and(path("/some/other/api")) + .respond_with(move |req: &wiremock::Request| { + *captured_clone.lock().unwrap() = Some(req.body.clone()); + ResponseTemplate::new(200).set_body_string("ok") + }) + .mount(&upstream) + .await; + + let proxy = start_proxy_with(&upstream.uri(), |c| { + c.compression = true; + }) + .await; + + // Same oversized JSON payload, but at a non-LLM path. The path + // gate must skip and the body must arrive verbatim. + let payload = oversized_anthropic_payload(); + let body = serde_json::to_vec(&payload).unwrap(); + let resp = reqwest::Client::new() + .post(format!("{}/some/other/api", proxy.url())) + .header("content-type", "application/json") + .body(body.clone()) + .send() + .await + .unwrap(); + assert_eq!(resp.status(), 200); + + let got = captured.lock().unwrap().clone().expect("upstream got body"); + assert_eq!(got, body, "non-LLM path must bypass compression"); + proxy.shutdown().await; +} diff --git a/plugins/headroom-agent-hooks/.claude-plugin/plugin.json b/plugins/headroom-agent-hooks/.claude-plugin/plugin.json index 8853264fa..779a2a0ba 100644 --- a/plugins/headroom-agent-hooks/.claude-plugin/plugin.json +++ b/plugins/headroom-agent-hooks/.claude-plugin/plugin.json @@ -1,6 +1,6 @@ { "name": "headroom", - "version": "0.20.2", + "version": "0.20.8", "description": "Headroom startup hooks for Claude Code and GitHub Copilot CLI.", "author": { "name": "Headroom Contributors", diff --git a/plugins/headroom-agent-hooks/.github/plugin/plugin.json b/plugins/headroom-agent-hooks/.github/plugin/plugin.json index efbb4aa63..3c948b9aa 100644 --- a/plugins/headroom-agent-hooks/.github/plugin/plugin.json +++ b/plugins/headroom-agent-hooks/.github/plugin/plugin.json @@ -1,6 +1,6 @@ { "name": "headroom", - "version": "0.20.2", + "version": "0.20.8", "description": "Headroom startup hooks for Claude Code and GitHub Copilot CLI.", "author": { "name": "Headroom Contributors", diff --git a/scripts/refresh_model_limits.sh b/scripts/refresh_model_limits.sh new file mode 100755 index 000000000..caf3a005d --- /dev/null +++ b/scripts/refresh_model_limits.sh @@ -0,0 +1,48 @@ +#!/usr/bin/env bash +# +# Refresh the vendored LiteLLM model_prices_and_context_window.json +# used by `crates/headroom-proxy/src/compression/model_limits.rs`. +# +# We vendor the snapshot rather than fetching at build/runtime so the +# proxy binary ships with no network dependency at startup. Operators +# tracking new model releases run this script and commit the diff. +# +# Validation: +# 1. JSON parses +# 2. Contains a known-stable Claude model entry +# 3. Contains a known-stable GPT model entry +# These guard against accidentally vendoring an empty / malformed file. + +set -euo pipefail + +REPO_ROOT="$(cd "$(dirname "$0")/.." && pwd)" +DEST="$REPO_ROOT/crates/headroom-proxy/data/model_prices_and_context_window.json" +URL="https://raw.githubusercontent.com/BerriAI/litellm/main/model_prices_and_context_window.json" + +echo "Fetching $URL" +TMP="$(mktemp)" +trap 'rm -f "$TMP"' EXIT +curl -fsSL "$URL" -o "$TMP" + +# Validate the snapshot before swapping it in. +python3 -c " +import json, sys +with open('$TMP') as f: + data = json.load(f) +if not isinstance(data, dict): + sys.exit('top-level not an object') +if 'sample_spec' not in data: + sys.exit('missing sample_spec entry — schema may have changed') +# Spot-check stable entries. +required = ['claude-sonnet-4-5-20250929', 'gpt-4o-mini', 'gpt-4-turbo'] +missing = [k for k in required if k not in data] +if missing: + sys.exit(f'missing required entries: {missing!r}') +print(f'OK: {len(data)} entries, including {required}') +" + +mv "$TMP" "$DEST" +trap - EXIT + +echo "Updated $DEST" +echo "Run 'cargo test -p headroom-proxy --lib compression::model_limits' to verify."