Reference
Rate limits
Limits protect the platform and your account. Normal integrations never reach them.
#Limits
| Scope | Limit | Notes |
|---|---|---|
| Per API key | 120 requests per minute | Token bucket: bursts of up to 120 requests, refilling at 2 per second. |
| Failed authentication, per IP address | 30 per minute | After that, every request from that IP gets 429 until the bucket refills, even with a valid key. |
#Rate limit headers
Every authenticated response tells you where you stand:
X-RateLimit-Limit: 120
X-RateLimit-Remaining: 118When you exceed the limit, you get 429 with a Retry-After header (seconds). Here, 130 parallel requests with one key returned 121 normal responses and 9 of these:
HTTP/1.1 429 Too Many Requests
Retry-After: 1
X-RateLimit-Limit: 120
X-RateLimit-Remaining: 0
X-Request-Id: req_9866e3acd9013bb6{
"error": {
"type": "rate_limit_error",
"code": "rate_limited",
"message": "Too many requests. Retry after the number of seconds in the Retry-After header.",
"request_id": "req_9866e3acd9013bb6"
}
}#Staying within the limits
- Use webhooks instead of polling. If you must poll, poll each open payment at most every few seconds, and stop once it's final.
- Back off on
429. Wait at leastRetry-Afterseconds. The SDKs do this automatically. - Page efficiently. Use
limit=100for reports and exports. - Fix authentication errors quickly. A misconfigured server that keeps sending a wrong key gets its whole IP address throttled.
- Use one key per integration, so a busy batch job can't slow down your checkout.
Need higher limits? Contact support with your expected volume.