generated: '2026-08-12'
method: searched
source: >-
  https://web.archive.org/web/20241015182444/https://helpcenter.integralplatform.com/article/reporting-api
  (IAS Help Center, "Reporting API", Jul 11, 2024) plus live probes of
  https://data.integralplatform.com on 2026-08-12
limit_count: 0
published: false
note: >-
  Integral Ad Science documents no rate limits, no quotas and no throttling
  behaviour for the Reporting API, and returns no rate-limit headers on the
  responses observable without credentials. An agent calling this API has no
  runtime signal telling it how hard it may poll — which matters more than usual
  here, because the API is job-oriented and the ONLY documented way to learn a
  report is finished is to poll GET /report-status in a loop with no documented
  interval.
limits: []
response_headers:
  observed: []
  ratelimit_standard: false
  x_ratelimit: false
  retry_after: false
  probe:
    url: https://data.integralplatform.com/report
    method: GET
    http_status: 401
    headers_returned:
      - date
      - content-length
      - cache-control
      - pragma
      - expires
      - strict-transport-security
      - x-frame-options
      - x-xss-protection
      - content-security-policy
      - referrer-policy
      - set-cookie
    note: no RateLimit-*, X-RateLimit-* or Retry-After header present
status_on_exhaustion: null
polling_guidance:
  documented: false
  recommended_interval: null
  note: >-
    IAS documents the SUBMITTED -> PENDING -> READY polling pattern but not a
    poll interval, a maximum poll rate, or a job-concurrency cap.
concurrency_limits:
  documented: false
report_size_limits:
  documented: false
  note: IAS documents metric minimum thresholds for statistical validity in the
    help center, but those are data-quality floors, not API quotas.