Rate limits
| API | Limit |
|---|---|
GraphQL (/graphql) |
600 requests a minute |
REST (/rest) |
600 requests a minute |
| Without authentication | 60 requests a minute, for each IP address |
The limit counts for each member, so all the API keys of one member share it. GraphQL and REST each have their own limit. The limit is a token bucket: it fills again over the minute, so short bursts are fine.
Headers
Section titled “Headers”Each answer tells you where you are:
| Header | Meaning |
|---|---|
X-RateLimit-Limit |
The size of the bucket. |
X-RateLimit-Remaining |
The requests that are left now. |
X-RateLimit-Reset |
The unix time when the bucket is full again. |
Over the limit
Section titled “Over the limit”A request over the limit gets 429 Too Many Requests with a Retry-After header, in seconds. Wait that long, then retry.
Some tips to stay below the limit:
- Ask for more records in one page (
itemsPerPage: 100). - Ask for the relations in the same GraphQL query, not in a separate request for each record.
- Use webhooks instead of polling for changes.