Skip to main content

Call Limits

In order to keep our service running a high level of performance, every account has a limit of requests. If you exceed your allotted limits, a 429 Too Many Requests error code will be returned and future requests may be throttled.

Each request returns a maximum of 200 results, use Pagination to fetch all your results with multiple requests. A limit above 200 is not an error — it is replaced by 200. See Filtering and Sorting for the full argument rules.

Response caching​

A successful response is cached for 10 minutes and replayed to you for that period. A record created or updated in the meantime will not appear until the cached copy expires, so a query polled in a tight loop returns the same payload each time.

The cache key is your account plus the exact query text, which has two practical consequences:

  • Two queries that differ only in whitespace are cached separately. Changing your query at all — even by a space — produces a fresh result rather than a cached one.
  • Only your account identifier and the query text form the key. Variables and the operation name sent alongside the query are not part of it, so two requests whose query text is identical share a cached response even when their variables differ.

Queries that return errors are not cached.