Rate limits
Bellwork API request limits and retry behavior
Bellwork applies per-organization request limits to protect the service.
| Plan | Requests per minute | Maximum page size |
|---|---|---|
| Free | 60 | 25 |
| Enterprise | 5,000 | 1,000 |
| Internal | 10,000 | 1,000 |
Individual endpoints can set a lower page-size maximum. For example, the signals feed currently accepts at most 200 rows per request.
When a request is rate-limited, Bellwork returns 429 Too Many Requests with
rate-limit headers describing the limit, remaining capacity, and reset time.
Successful responses may also carry those headers.
Use bounded exponential backoff with jitter for 429 and transient 5xx
responses. Do not retry validation (400), authentication (401), entitlement
(403), or not-found (404) responses without changing the request.
For large datasets, page sequentially using the endpoint's limit and
offset parameters. Stop when pagination.has_more or the feed's has_more
is false.