generated: '2026-07-20'
method: searched
source: https://apidocs.persistiq.com/
scope: per-api-key
limits:
  - name: default
    limit_count: 100
    window: 1 minute
    applies_to: all endpoints
    response_status: 429
headers:
  - name: X-RateLimit-Limit
    description: Maximum allowed requests per minute.
  - name: X-RateLimit-Remaining
    description: Number of requests left for the current window.
  - name: X-RateLimit-Reset
    description: UTC epoch seconds at which the current window ends and the limit resets.
over_limit:
  status: 429
  reason: too_many_requests
  body: >-
    { "error": { "reason": "too_many_requests", "message": "You performed more
    than 100 requests per minute. Limit will be reseted at <epoch>." } }
observed:
  method: probed
  fetched: '2026-08-13'
  request: GET https://api.persistiq.com/v1/users (no api key)
  http_status: 401
  headers:
    x-ratelimit-limit: '500'
    x-ratelimit-remaining: '498'
    x-request-id: present (UUID)
    x-runtime: present
    strict-transport-security: max-age=31536000
  notes: >-
    Rate-limit headers are returned even on an unauthenticated 401, so an agent
    can read the live budget before it holds a key. No `x-ratelimit-reset` and
    no `Retry-After` appeared on this response; the reference documents a reset
    header, so it is likely emitted only as the window is consumed. Header names
    are returned lower-cased over HTTP/2.
discrepancy:
  documented: 100 requests / 1 minute (https://apidocs.persistiq.com/)
  observed: 500 (x-ratelimit-limit, 2026-08-13)
  resolution: unresolved
  guidance: >-
    Treat the runtime headers as authoritative and the documented 100/min as the
    floor. The gap is recorded rather than reconciled — PersistIQ has not
    restated the limit anywhere machine-readable, and its official OpenAPI does
    not describe rate limiting at all.
notes: >-
  A maximum of 100 requests is allowed in any 1-minute window, enforced per API
  key (i.e. per user), per the published reference. Exceeding the limit returns
  429 Too Many Requests. A live probe on 2026-08-13 advertised a limit of 500 —
  see `observed` and `discrepancy` above.