generated: '2026-08-12'
method: searched
source:
- https://platform.cardlytics.com/advertisers/docs/api-merchant-rest-api
- https://platform.cardlytics.com/advertisers/docs/api-faqs
- https://docs.cardlytics.com/ads/v2/error-handling/index.html
- https://docs.cardlytics.com/poweredby/api-reference-authentication.html
limit_count: 0
note: >-
  Cardlytics documents that rate limits EXIST and that exceeding them returns
  429, but publishes no numeric limit any caller can rely on. Partner API quotas
  ("a daily quota of the number of calls" and "an RPS, for example, 10 calls per
  second") are provisioned per partner in the partner agreement — the "10 calls
  per second" in the docs is explicitly an example, not a published ceiling, so
  it is not recorded as a limit here. On the Publisher API the provider states
  the opposite posture: "Cardlytics imposes no general rules around throttling
  incoming requests" and "tries to never respond with this response code".
  limit_count is therefore 0: an honest zero, not an unchecked field.
limits:
- scope: per-partner (Partner API, advertiser merchant/offer ingestion)
  window: day
  limit: null
  provisioned: true
  note: >-
    "Every Partner will be provided with a daily quota of the number of calls
    they can make, this should be ~ Number of merchant offers." The value is
    per-partner and not published.
  status_on_exhaustion: 429
  source: https://platform.cardlytics.com/advertisers/docs/api-merchant-rest-api
- scope: per-partner (Partner API)
  window: second
  limit: null
  provisioned: true
  note: >-
    "Every Partner will be provided with an RPS, for example, 10 calls per
    second." The example value is illustrative; the real RPS is assigned per
    partner.
  status_on_exhaustion: 429
  source: https://platform.cardlytics.com/advertisers/docs/api-merchant-rest-api
- scope: per-request payload cap (Partner API aggregate reporting)
  window: per-request
  limit: 1000
  unit: offer IDs per report request
  source: https://platform.cardlytics.com/advertisers/docs/api-faqs
- scope: global (Publisher API v2, Powered by Cardlytics)
  window: null
  limit: null
  note: >-
    No general throttling rules imposed; publishers are nevertheless asked to
    honor 429 with exponential backoff.
  status_on_exhaustion: 429
  source: https://docs.cardlytics.com/ads/v2/error-handling/index.html
response_headers: []
response_headers_note: >-
  No X-RateLimit-*, RateLimit-* or Retry-After response headers are documented on
  any Cardlytics API, and none appear in any of the three published OpenAPI
  documents. An agent gets no runtime budget signal — only a 429 after the fact.
status_on_exhaustion: 429
retry_guidance:
  strategy: exponential backoff
  documented: true
  source: https://docs.cardlytics.com/ads/v2/error-handling/index.html
header_size_limits:
- header: X-CDLX-Session
  max_bytes: 512
  behavior: silently truncated past 512 bytes
  source: https://docs.cardlytics.com/ads/v1/common/custom-http-headers.html
- header: X-CDLX-Meta
  max_bytes: 512
  behavior: silently truncated past 512 bytes
  source: https://docs.cardlytics.com/ads/v1/common/custom-http-headers.html