generated: '2026-07-21'
method: searched
source: https://run-ai-docs.nvidia.com/api/getting-started/using-the-rest-api/making-rest-api-requests.md
docs:
- https://run-ai-docs.nvidia.com/api/getting-started/about-the-rest-api.md
- https://run-ai-docs.nvidia.com/api/getting-started/using-the-rest-api/pagination.md
- https://run-ai-docs.nvidia.com/api/getting-started/how-to-authenticate-to-the-api.md
authentication:
  style: bearer-jwt
  scheme: Authorization Bearer <accessToken>
  token_endpoint: POST https://<COMPANY-URL>/api/v1/token
  grant_type: client_credentials
  credential: access key (clientId + clientSecret) for a user or service account
  ref: authentication/runai-authentication.yml
content_type:
  request: application/json
  response: application/json
  methods: [GET, POST, PUT, PATCH, DELETE]
idempotency:
  supported: false
  note: >-
    No Idempotency-Key header or idempotency contract is documented or present in
    the OpenAPI. Not emitting an Idempotency pointer.
pagination:
  style: offset-limit
  params:
    limit: Maximum items to return (default 50, range 1..500)
    offset: Offset of the first item returned
  response_fields:
    collection: named after the resource (e.g. workloads, projects)
    next: pointer/offset for the next page; absent or empty when the end is reached
  example: GET /api/v1/workloads?offset=0&limit=50
filtering:
  sorting:
    params: [sortBy, sortOrder]
  structured_filter:
    param: filterBy
    operators: ['==', '!=', '>=', '=@ (contains)']
    example: filterBy=allocatedGPU>=2
  free_text:
    param: search
versioning:
  scheme: dual-track (latest continuously updated; versioned self-hosted releases 2.x)
  ref: lifecycle/runai-lifecycle.yml
error_envelope:
  format: json-envelope
  shape: '{code:int, message:string, details?:string}'
  ref: errors/runai-problem-types.yml
rate_limit_signaling:
  documented: false
  note: No rate-limit headers or policy documented in the OpenAPI or docs.