generated: '2026-08-13'
method: searched
source:
- https://open.thetradedesk.com/advertiser/docsApp/Foundations/resources/doc/RateLimits
- https://open.thetradedesk.com/advertiser/docsApp/Foundations/resources/doc/GqlApiRateLimits
- https://open.thetradedesk.com/advertiser/docsApp/Foundations/resources/doc/ReturnCodes
limit_count: 0
summary: The Trade Desk enforces per-endpoint, per-user-account rate limits but publishes NO numeric limit
  for any endpoint. The docs describe the mechanism, the response signal and the mitigation strategy in
  detail, then route every "what is my actual limit" question to a Technical Account Manager. limit_count
  is 0 because zero numeric limits are published — that is the finding, not a gap in this harvest.
limits: []
mechanism:
  scope:
  - per-endpoint
  - per-user-account
  - per-operation (GraphQL)
  - group throttling across multiple user accounts under one client
  window: usually one minute
  dynamic: System-level usage is monitored and can trigger further dynamic reductions to protect the platform.
    High-cost GraphQL requests may be dynamically adjusted during periods of high demand.
  ui_counts: Platform UI usage counts against the same limits — the UI calls the API with the same account
    credentials.
  rest_vs_graphql: Limits are enforced the same way for REST and GraphQL, but the REST call count and
    the GraphQL call count are independent of each other.
  increase: Contact your Technical Account Manager.
response_signals:
  status_code: 429
  status_description: Too Many Requests
  response_headers:
  - header: Retry-After
    description: Seconds to wait before retrying. The provider explicitly instructs clients to inspect
      it and extract the value in seconds.
    documented: true
  headers_absent:
  - X-RateLimit-Limit
  - X-RateLimit-Remaining
  - X-RateLimit-Reset
  - RateLimit-Limit
  - RateLimit-Remaining
  - RateLimit-Reset
  headers_absent_note: No quota-disclosure headers are documented. An agent can learn it has been throttled,
    and for how long, but never how much budget it has left.
  also:
  - status: 503
    header: Retry-After
    description: Service Unavailable — the docs direct clients to use Retry-After as guidance on when
      to retry.
client_guidance:
- Limit concurrency to four callers per endpoint (REST) or per operation (GraphQL).
- 'Fixed delay: wait one minute after a failed call before retrying.'
- 'Exponential backoff: Delay = Min(max_delay, base_delay * 2 ^ retrycount).'
- Cache data that does not change frequently (industry categories, first/third-party DMP lookups, geosegment
  queries).
- Use delta endpoints and delta mutations to reduce execution time and call frequency.
- Batch GraphQL mutations rather than issuing them one at a time.
- 'On POST /v3/dmp/thirdparty/advertiser set ExcludeTotalCounts: true, scope with BrandIds or ThirdPartyDataIds,
  and keep PageSize at or below 1000.'
graphql_complexity:
  documented: true
  docs: https://open.thetradedesk.com/advertiser/docsApp/Foundations/resources/doc/GqlApiErrors
  note: GraphQL additionally enforces query complexity limits on nested queries, documented separately
    from rate limits.
request_size_limits:
- api: Real-Time Conversion Events
  limit: 28.6 MB per request
  note: No limit on the number of events per request.
  source: https://open.thetradedesk.com/advertiser/docsApp/GuidesAdvertiser/data/doc/post-track-realtimeconversion
- api: Data API
  limit: unpublished
  signal: HTTP 413 with error code MaxRequestSizeExceeded