NVIDIA NIM

aid: nvidia-nim
url: https://raw.githubusercontent.com/api-evangelist/nvidia-nim/refs/heads/main/apis.yml
apis:
  - aid: nvidia-nim:nvidia-nim-chat-completions-api
    name: NVIDIA NIM Chat Completions API
    tags:
      - AI
      - Artificial Intelligence
      - Chat
      - Completions
      - LLM
      - OpenAI Compatible
    humanURL: https://docs.api.nvidia.com/nim/reference/llm-apis
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.api.nvidia.com/nim/reference/llm-apis
        type: Documentation
      - url: https://docs.nvidia.com/nim/large-language-models/latest/api-reference.html
        type: Documentation
      - url: openapi/nvidia-nim-chat-completions-api-openapi.yml
        type: OpenAPI
      - url: json-schema/nvidia-nim-chat-completion-schema.json
        type: JSONSchema
      - url: json-ld/nvidia-nim-context.jsonld
        type: JSONLD
    description: >-
      OpenAI-compatible chat completions endpoint exposing 100+ foundation models — Meta Llama, Mistral, Mixtral, NVIDIA
      Nemotron, DeepSeek, Qwen, Microsoft Phi, Google Gemma, IBM Granite, and more — through a single
      /v1/chat/completions surface. Supports streaming, tool/function calling, structured outputs, vision inputs on
      multimodal models, and the standard temperature/top_p/max_tokens parameters. Switching models is a one-line change
      to the model string. Available hosted on integrate.api.nvidia.com or self-hosted via NIM containers on any GPU.
  - aid: nvidia-nim:nvidia-nim-completions-api
    name: NVIDIA NIM Completions API
    tags:
      - AI
      - Artificial Intelligence
      - Completions
      - LLM
      - OpenAI Compatible
    humanURL: https://docs.nvidia.com/nim/large-language-models/latest/api-reference.html
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.nvidia.com/nim/large-language-models/latest/api-reference.html
        type: Documentation
      - url: openapi/nvidia-nim-completions-api-openapi.yml
        type: OpenAPI
    description: >-
      Legacy OpenAI-compatible text completion endpoint (/v1/completions) for non-chat foundation models served by NIM.
      Accepts a raw prompt and returns generated text with the same streaming, sampling, and stopping-criterion controls
      as the chat endpoint.
  - aid: nvidia-nim:nvidia-nim-embeddings-api
    name: NVIDIA NIM Embeddings API
    tags:
      - AI
      - Artificial Intelligence
      - Embeddings
      - Retrieval
      - RAG
      - OpenAI Compatible
    humanURL: https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/api-reference.html
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.nvidia.com/nim/nemo-retriever/text-embedding/latest/api-reference.html
        type: Documentation
      - url: openapi/nvidia-nim-embeddings-api-openapi.yml
        type: OpenAPI
      - url: json-schema/nvidia-nim-embedding-schema.json
        type: JSONSchema
    description: >-
      OpenAI-compatible embeddings endpoint (/v1/embeddings) backed by NVIDIA NeMo Retriever text embedding models
      including NV-Embed, NV-EmbedQA-E5, llama-3.2-nv-embedqa-1b, and BAAI BGE-M3. Returns dense float vectors for
      documents or queries to power RAG, semantic search, and clustering. Supports `input_type=passage|query` for
      asymmetric retrieval and the standard `dimensions` parameter on models that permit dimension reduction.
  - aid: nvidia-nim:nvidia-nim-reranking-api
    name: NVIDIA NIM Reranking API
    tags:
      - AI
      - Artificial Intelligence
      - Reranking
      - Retrieval
      - RAG
    humanURL: https://docs.nvidia.com/nim/nemo-retriever/text-reranking/latest/api-reference.html
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.nvidia.com/nim/nemo-retriever/text-reranking/latest/api-reference.html
        type: Documentation
      - url: openapi/nvidia-nim-reranking-api-openapi.yml
        type: OpenAPI
    description: >-
      NeMo Retriever cross-encoder reranking endpoint (/v1/ranking) for scoring candidate passages against a query.
      Improves retrieval relevance on RAG pipelines and supports the llama-3.2-nv-rerankqa-1b and
      NV-RerankQA-Mistral-4B-v3 models. Accepts a query plus a list of passages and returns a sorted list of relevance
      scores.
  - aid: nvidia-nim:nvidia-nim-models-api
    name: NVIDIA NIM Models API
    tags:
      - AI
      - Artificial Intelligence
      - Models
      - Catalog
      - OpenAI Compatible
    humanURL: https://docs.nvidia.com/nim/large-language-models/latest/api-reference.html
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.nvidia.com/nim/large-language-models/latest/api-reference.html
        type: Documentation
      - url: openapi/nvidia-nim-models-api-openapi.yml
        type: OpenAPI
    description: >-
      OpenAI-compatible model catalog endpoint (/v1/models) returning the list of models served by the NIM endpoint or
      container. Each entry includes id, owned_by, and created timestamp. Used by clients to discover the model name
      strings to pass to chat-completions / completions / embeddings.
  - aid: nvidia-nim:nvidia-nim-vision-api
    name: NVIDIA NIM Vision Language Models API
    tags:
      - AI
      - Artificial Intelligence
      - Vision
      - Multimodal
      - VLM
    humanURL: https://docs.api.nvidia.com/nim/reference/vlm-apis
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.api.nvidia.com/nim/reference/vlm-apis
        type: Documentation
      - url: openapi/nvidia-nim-vision-api-openapi.yml
        type: OpenAPI
    description: >-
      Vision-language model inference through the standard /v1/chat/completions surface with image inputs (base64 or
      URL) in the messages payload. Supports NVIDIA NeVA, microsoft/kosmos-2, Phi-3-vision,
      llama-3.2-90b-vision-instruct, and other VLMs hosted in the NIM catalog.
  - aid: nvidia-nim:nvidia-nim-health-api
    name: NVIDIA NIM Health API
    tags:
      - Health
      - Observability
      - Kubernetes
    humanURL: https://docs.nvidia.com/nim/large-language-models/latest/observability.html
    properties:
      - url: https://docs.nvidia.com/nim/large-language-models/latest/observability.html
        type: Documentation
      - url: openapi/nvidia-nim-health-api-openapi.yml
        type: OpenAPI
    description: >-
      Liveness, readiness, and startup probes exposed by self-hosted NIM containers (/v1/health/live, /v1/health/ready)
      and a Prometheus /v1/metrics scrape endpoint for GPU utilization, request latency, and queue depth. Drives
      Kubernetes pod lifecycle and HPA scaling via the NIM Operator.
  - aid: nvidia-nim:nvidia-nim-image-generation-api
    name: NVIDIA NIM Image Generation API
    tags:
      - AI
      - Artificial Intelligence
      - Image Generation
      - Visual
    humanURL: https://docs.api.nvidia.com/nim/reference/visual-models
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.api.nvidia.com/nim/reference/visual-models
        type: Documentation
      - url: openapi/nvidia-nim-image-generation-api-openapi.yml
        type: OpenAPI
    description: >-
      Visual generative AI endpoints for text-to-image, image-to-image, and image editing using models such as Black
      Forest Labs FLUX.1, Stable Diffusion XL, Shutterstock-trained models, and NVIDIA-curated Edify Image. Returns
      base64-encoded PNG/JPEG artifacts.
  - aid: nvidia-nim:nvidia-nim-speech-api
    name: NVIDIA NIM Speech API
    tags:
      - AI
      - Artificial Intelligence
      - Speech
      - ASR
      - TTS
      - Riva
    humanURL: https://docs.nvidia.com/nim/riva/latest/index.html
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.nvidia.com/nim/riva/latest/index.html
        type: Documentation
      - url: openapi/nvidia-nim-speech-api-openapi.yml
        type: OpenAPI
    description: >-
      NVIDIA Riva-powered speech NIMs delivering automatic speech recognition (Parakeet, Canary), neural machine
      translation, and text-to-speech (Magpie-TTS, FastPitch) through HTTP and gRPC surfaces. Hosted endpoints accept
      WAV/FLAC audio and return transcripts or synthesized speech.
  - aid: nvidia-nim:nvidia-nim-biology-api
    name: NVIDIA NIM Biology (BioNeMo) API
    tags:
      - AI
      - Biology
      - BioNeMo
      - Drug Discovery
      - Healthcare
    humanURL: https://docs.nvidia.com/nim/bionemo/latest/index.html
    baseURL: https://integrate.api.nvidia.com/v1
    properties:
      - url: https://docs.nvidia.com/nim/bionemo/latest/index.html
        type: Documentation
      - url: openapi/nvidia-nim-biology-api-openapi.yml
        type: OpenAPI
    description: >-
      BioNeMo NIMs for protein structure prediction (AlphaFold2, ESMFold, OpenFold), protein generation (ProtGPT2,
      RFDiffusion), molecular property prediction (MolMIM), small molecule generation, and molecular docking (DiffDock).
      Each model is a containerized microservice with the same OpenAPI surface.
name: NVIDIA NIM
tags:
  - AI
  - Artificial Intelligence
  - Inference
  - Microservices
  - LLM
  - Foundation Models
  - GPU
  - Kubernetes
  - NVIDIA
  - OpenAI Compatible
kind: contract
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
access: 3rd-Party
common:
  - type: PostmanWorkspace
    url: https://www.postman.com/kinlaneapi/nvidia-nim/overview
  - type: ArazzoWorkflows
    url: arazzo/
    workflows:
      - url: arazzo/nvidia-nim-bionemo-drug-discovery-workflow.yml
        name: NVIDIA NIM BioNeMo Drug Discovery
        summary: >-
          Fold a protein from sequence, dock a ligand into the predicted structure, then generate optimized analog
          molecules.
      - url: arazzo/nvidia-nim-discover-and-chat-workflow.yml
        name: NVIDIA NIM Discover And Chat
        summary: List the served models, confirm a target model's metadata, then run a chat completion against it.
      - url: arazzo/nvidia-nim-generate-image-and-caption-workflow.yml
        name: NVIDIA NIM Generate Image And Caption
        summary: Generate an image from a text prompt, then caption the generated image with a vision-language model.
      - url: arazzo/nvidia-nim-health-gated-completion-workflow.yml
        name: NVIDIA NIM Health Gated Completion
        summary: Check a self-hosted NIM container's readiness, and only run a text completion once the engine reports ready.
      - url: arazzo/nvidia-nim-rag-rerank-answer-workflow.yml
        name: NVIDIA NIM RAG Rerank And Answer
        summary: Embed a query, rerank candidate passages against it, then answer the question grounded in the top passage.
      - url: arazzo/nvidia-nim-vision-describe-and-summarize-workflow.yml
        name: NVIDIA NIM Vision Describe And Summarize
        summary: >-
          Describe an image with a vision-language model, then condense the description into a short caption with an
          LLM.
      - url: arazzo/nvidia-nim-voice-assistant-loop-workflow.yml
        name: NVIDIA NIM Voice Assistant Loop
        summary: >-
          Transcribe an audio clip with Riva ASR, answer the transcript with an LLM, then synthesize the reply with Riva
          TTS.
  - type: Portal
    url: https://build.nvidia.com
  - type: Documentation
    url: https://docs.nvidia.com/nim/index.html
  - type: Documentation
    url: https://docs.api.nvidia.com/nim/reference/llm-apis
  - type: Documentation
    url: https://developer.nvidia.com/nim
  - type: GettingStarted
    url: https://docs.nvidia.com/nim/large-language-models/latest/getting-started.html
  - type: SignUp
    url: https://build.nvidia.com/explore/discover
  - type: Sandbox
    url: https://build.nvidia.com/explore/discover
  - type: Pricing
    url: https://www.nvidia.com/en-us/data-center/products/ai-enterprise/
  - type: GitHubOrganization
    url: https://github.com/NVIDIA
  - type: GitHubOrganization
    url: https://github.com/NVIDIA-NIM-Agent-Blueprints
  - type: StatusPage
    url: https://status.nvidia.com
  - type: Blog
    url: https://developer.nvidia.com/blog/category/generative-ai/
  - type: Blog
    url: https://blogs.nvidia.com/blog/category/generative-ai/
  - type: Forum
    url: https://forums.developer.nvidia.com/c/ai-data-science/nemo-llm-service/
  - type: TrustCenter
    url: https://www.nvidia.com/en-us/about-nvidia/legal-info/
  - type: TermsOfService
    url: https://www.nvidia.com/en-us/about-nvidia/terms-of-service/
  - type: PrivacyPolicy
    url: https://www.nvidia.com/en-us/about-nvidia/privacy-policy/
  - type: Documentation
    url: https://docs.nvidia.com/nim-operator/latest/index.html
    name: NIM Operator Documentation
  - type: SDK
    url: https://github.com/NVIDIA/nim-deploy
    name: NIM Deploy (Helm Charts and Reference Implementations)
  - type: SDK
    url: https://github.com/NVIDIA/k8s-nim-operator
    name: Kubernetes NIM Operator
  - type: SDK
    url: https://github.com/NVIDIA/GenerativeAIExamples
    name: Generative AI Examples
  - type: SDK
    url: https://github.com/NVIDIA/NeMo
    name: NeMo Toolkit
  - type: SDK
    url: https://github.com/NVIDIA/NeMo-Guardrails
    name: NeMo Guardrails
  - type: SDK
    url: https://github.com/NVIDIA/TensorRT-LLM
    name: TensorRT-LLM
  - type: SDK
    url: https://github.com/triton-inference-server/server
    name: Triton Inference Server
  - type: SDK
    url: https://github.com/langchain-ai/langchain-nvidia
    name: LangChain NVIDIA AI Endpoints
  - type: SDK
    url: https://pypi.org/project/openai/
    name: OpenAI Python SDK (compatible)
  - type: CodeExamples
    url: https://github.com/NVIDIA/GenerativeAIExamples
    name: NVIDIA Generative AI Examples
  - type: CodeExamples
    url: https://github.com/NVIDIA-AI-Blueprints
    name: NVIDIA AI Blueprints
  - type: Models
    url: https://build.nvidia.com/explore/discover
  - type: KubernetesCRD
    url: https://github.com/NVIDIA/k8s-nim-operator/tree/main/api
    name: NIMService / NIMCache / NIMPipeline CRDs
  - type: RateLimits
    url: https://docs.api.nvidia.com/nim/reference/limits
  - type: Versioning
    url: https://docs.nvidia.com/nim/large-language-models/latest/release-notes.html
  - url: plans/nvidia-nim-plans-pricing.yml
    type: Plans
  - url: rate-limits/nvidia-nim-rate-limits.yml
    type: RateLimits
  - url: finops/nvidia-nim-finops.yml
    type: FinOps
  - type: Features
    data:
      - OpenAI-compatible REST surface — /v1/chat/completions, /v1/completions, /v1/embeddings, /v1/models, /v1/ranking
      - >-
        100+ foundation models exposed through a single API contract — Llama 3.1/3.2/3.3, Mistral, Mixtral, NVIDIA
        Nemotron, DeepSeek-R1, Qwen 2.5, Microsoft Phi, Google Gemma, IBM Granite, and Falcon
      - Free hosted inference at build.nvidia.com on DGX Cloud — 1,000 credits on signup, 40 RPM rate limit
      - Self-hosted deployment via Docker containers shipping TensorRT-LLM, vLLM, or SGLang inference engines
      - Kubernetes-native deployment via the NIM Operator with NIMService, NIMCache, NIMPipeline CRDs
      - GPU-aware autoscaling, persistent model caches, and rolling upgrades managed by the operator
      - Multi-tenant licensing through NVIDIA AI Enterprise (commercial production use)
      - NeMo Retriever NIMs for embeddings, reranking, OCR, and PDF-to-Markdown extraction in RAG pipelines
      - Vision Language Model NIMs reusing the chat-completions surface for multimodal inputs
      - NVIDIA Riva speech NIMs (Parakeet ASR, Canary translation, Magpie TTS) with HTTP and gRPC adapters
      - BioNeMo NIMs for AlphaFold2, ESMFold, ProtGPT2, MolMIM, DiffDock, RFDiffusion
      - Visual generative AI NIMs — FLUX.1, SDXL, Shutterstock Edify Image, Edify 3D
      - NeMo Guardrails for input/output safety and topic policy enforcement
      - Function calling, JSON mode, tool use, and structured outputs across compatible LLMs
      - Streaming via Server-Sent Events on chat/completions
      - Prometheus /v1/metrics scrape endpoint and /v1/health/{live,ready} probes for Kubernetes
      - LangChain, LlamaIndex, Haystack, OpenAI SDK, and direct REST client compatibility
      - NVIDIA AI Blueprints — full reference RAG, multimodal search, drug discovery, and digital human stacks
      - Available on DGX Cloud, AWS, Azure, Google Cloud, Oracle Cloud, GKE, EKS, AKS, OpenShift, and on-prem
    sources:
      - https://build.nvidia.com
      - https://docs.nvidia.com/nim/index.html
      - https://docs.api.nvidia.com/nim/reference/llm-apis
      - https://www.nvidia.com/en-us/ai-data-science/products/nim-microservices/
      - https://github.com/NVIDIA/k8s-nim-operator
      - https://github.com/NVIDIA/nim-deploy
    updated: '2026-05-25'
created: '2026-05-25'
modified: '2026-05-25'
position: Consuming
description: >-
  NVIDIA NIM (NVIDIA Inference Microservices) is a catalog of GPU-accelerated, containerized AI inference microservices
  that package optimized model engines (TensorRT-LLM, vLLM, SGLang, Triton) behind industry-standard OpenAI-compatible
  REST APIs. NIM covers large language models, embeddings and reranking, vision-language models, speech (Riva), visual
  generative AI, and biology (BioNeMo) — exposed identically whether consumed from the hosted endpoint at
  integrate.api.nvidia.com or self-hosted via Docker containers and the Kubernetes-native NIM Operator. NIM ships with
  NVIDIA AI Enterprise for commercial deployment and is the inference layer underneath NVIDIA AI Blueprints, NeMo
  Retriever, NeMo Guardrails, and the broader NVIDIA developer stack.
maintainers:
  - FN: Kin Lane
    email: info@apievangelist.com
    X: apievangelist
    url: https://apievangelist.com
specificationVersion: '0.16'