generated: '2026-09-18'
method: searched
source: https://apidocs.myreliaquest.com/ (GreyMatter API documentation, "Rate Limiting" section, fetched 2026-09-18)
api: reliaquest:greymatter-api
limit_count: 1
scopes:
  - scope: per-user
    surface: GraphQL (https://greymatter.myreliaquest.com/graphql)
    window: 1 hour
    limit: 5000
    unit: tokens
    burst: null
    counted: >-
      "Each Node entity in the request counts as a token" — the cost is the number of Node entities
      a query touches, not the number of HTTP requests, so a wide paginated query spends its
      allowance faster than a narrow one.
    status_on_exhaustion: null
    status_note: >-
      The docs say only that 4xx codes indicate a request problem; the exact status returned on
      exhaustion (429 or otherwise) is not stated and was not observed — the endpoint answers 403 to
      unauthenticated callers.
    response_headers: []
    headers_note: >-
      No rate-limit response headers (X-RateLimit-*, RateLimit-*, Retry-After) are documented. The
      runtime signal ReliaQuest provides instead is an in-band GraphQL query, `rateLimit { cost }`
      (Utilities folder, returns CostEstimationResponse), which reports the token cost — an agent
      should poll that rather than expect a header.
    cost_query:
      operation: rateLimit
      document: 'query rateLimit { rateLimit { cost } }'
      returns: CostEstimationResponse
searchlight_note: >-
  No rate limits are published for the Digital Shadows SearchLight Portal API; its reference is
  customer-only. Not counted above.