generated: '2026-06-20'
method: derived
status: candidate
source: openapi/*-openapi.yml
notes: NVIDIA publishes no first-party hosted/remote MCP server for the NIM inference API. NVIDIA's MCP
  story lives in the NeMo Agent Toolkit, which can run as a self-hosted MCP server (streamable-http, default
  http://localhost:9901/mcp) that publishes workflow tools and can also act as an MCP client — but that
  is an agent-framework server, not a hosted endpoint that fronts the NIM REST surface. There is also
  a community NVIDIA NIM MCP server (not first-party). The tool list below is DERIVED one-per-operation
  from the repo's OpenAPI as a candidate; it is not a published NVIDIA server.
references:
- name: NeMo Agent Toolkit MCP server
  url: https://docs.nvidia.com/nemo/agent-toolkit/latest/run-workflows/mcp-server.html
  first_party: true
  hosted: false
server:
  name: nvidia-nim
  transport: http
tools:
- name: create_chat_completion
  description: Generate an OpenAI-compatible chat completion across 100+ foundation models.
  source_operation: openapi/nvidia-nim-chat-completions-api-openapi.yml#createChatCompletion
- name: create_completion
  description: Generate a legacy text completion from a raw prompt.
  source_operation: openapi/nvidia-nim-completions-api-openapi.yml#createCompletion
- name: create_embedding
  description: Return dense float embedding vectors for documents or queries.
  source_operation: openapi/nvidia-nim-embeddings-api-openapi.yml#createEmbedding
- name: rank_passages
  description: Rerank candidate passages against a query with a cross-encoder.
  source_operation: openapi/nvidia-nim-reranking-api-openapi.yml#rankPassages
- name: list_models
  description: List models served by the endpoint or container.
  source_operation: openapi/nvidia-nim-models-api-openapi.yml#listModels
- name: get_model
  description: Retrieve a single model's metadata.
  source_operation: openapi/nvidia-nim-models-api-openapi.yml#getModel
- name: create_vision_chat_completion
  description: Multimodal chat completion with image inputs on VLM models.
  source_operation: openapi/nvidia-nim-vision-api-openapi.yml#createVisionChatCompletion
- name: generate_image
  description: Text-to-image / image-to-image visual generation.
  source_operation: openapi/nvidia-nim-image-generation-api-openapi.yml#generateImage
- name: create_transcription
  description: Transcribe audio to text via Riva ASR NIMs.
  source_operation: openapi/nvidia-nim-speech-api-openapi.yml#createTranscription
- name: create_speech
  description: Synthesize speech from text via Riva TTS NIMs.
  source_operation: openapi/nvidia-nim-speech-api-openapi.yml#createSpeech
- name: predict_protein_structure
  description: Predict protein structure via BioNeMo (AlphaFold2/ESMFold/OpenFold).
  source_operation: openapi/nvidia-nim-biology-api-openapi.yml#predictProteinStructure
- name: dock_ligand
  description: Molecular docking of a ligand to a protein (DiffDock).
  source_operation: openapi/nvidia-nim-biology-api-openapi.yml#dockLigand
- name: generate_molecules
  description: Generate small molecules (MolMIM).
  source_operation: openapi/nvidia-nim-biology-api-openapi.yml#generateMolecules
- name: get_liveness
  description: Container liveness probe.
  source_operation: openapi/nvidia-nim-health-api-openapi.yml#getLiveness
- name: get_readiness
  description: Container readiness probe.
  source_operation: openapi/nvidia-nim-health-api-openapi.yml#getReadiness
deployment:
  mode: none
  verified: derived
  tools: 15
  checked: '2026-08-12'
  source: catalog MCP census