generated: '2026-06-20'
method: derived
source: openapi/*-openapi.yml
notes: >-
  DERIVED entity graph from the repo OpenAPI. NIM is a stateless inference surface: there are
  no persisted, cross-referencing resources like a CRM/payments API. The only first-class,
  retrievable resource is Model (/v1/models); every inference request references a Model by its
  `model` id string. Request/response payloads (ChatCompletion, Embedding, Ranking, Image,
  Transcription, Speech, StructurePrediction) are ephemeral generations keyed to a Model.
entities:
  - name: Model
    retrievable: true
    id_field: id
    endpoints: [/v1/models, '/v1/models/{model}']
    example_ids: [meta/llama-3.3-70b-instruct, nvidia/llama-3.1-nemotron-70b-instruct, black-forest-labs/flux.1-dev]
    fields: [id, object, created, owned_by]
  - name: ChatCompletion
    retrievable: false
    ephemeral: true
    endpoint: /v1/chat/completions
  - name: Completion
    retrievable: false
    ephemeral: true
    endpoint: /v1/completions
  - name: Embedding
    retrievable: false
    ephemeral: true
    endpoint: /v1/embeddings
  - name: Ranking
    retrievable: false
    ephemeral: true
    endpoint: /v1/ranking
  - name: ImageGeneration
    retrievable: false
    ephemeral: true
    endpoint: /v1/images/generations
  - name: Transcription
    retrievable: false
    ephemeral: true
    endpoint: /v1/audio/transcriptions
  - name: Speech
    retrievable: false
    ephemeral: true
    endpoint: /v1/audio/speech
  - name: StructurePrediction
    retrievable: false
    ephemeral: true
    endpoint: /v1/biology/*
relationships:
  - from: ChatCompletion
    to: Model
    type: belongs_to
    via: model
  - from: Completion
    to: Model
    type: belongs_to
    via: model
  - from: Embedding
    to: Model
    type: belongs_to
    via: model
  - from: Ranking
    to: Model
    type: belongs_to
    via: model
  - from: ImageGeneration
    to: Model
    type: belongs_to
    via: model
  - from: Transcription
    to: Model
    type: belongs_to
    via: model
  - from: Speech
    to: Model
    type: belongs_to
    via: model
  - from: StructurePrediction
    to: Model
    type: belongs_to
    via: model