API docs · Reliability
Rate limits
RPM, concurrency and rate-limit response headers.
RPM is counted per key over a fixed 60-second window. The limit is the lower of the key's RPM and the account's RPM, and each key in an account is counted separately. Requests over the limit return 429 rate_limited. Every authenticated request's response includes these headers:
| Header | Meaning |
|---|---|
X-RateLimit-Limit | Requests allowed in the current window |
X-RateLimit-Remaining | Requests remaining in the current window |
X-RateLimit-Reset | When the window resets (unix seconds) |
Retry-After | Only on 429: suggested number of seconds to wait (10 when the concurrency limit is exceeded) |
The number of tasks in progress at once is capped by both the key's and the account's concurrency limit. You can set RPM and concurrency limits for each key in the console; contact our operations team if you need higher account limits.
429 response
HTTP/1.1 429 Too Many Requests
Retry-After: 12
X-RateLimit-Limit: 60
X-RateLimit-Remaining: 0
X-RateLimit-Reset: 1730259012
{ "error": { "code": "rate_limited", "retryable": true, ... } }