Rate Limits
Respect per-key limits and rate-limit response headers.
Rate limits apply per API key across all public endpoints.
Limits
| Environment | Limit | Window |
|---|---|---|
| Test Sandbox | 60 requests | 60 seconds |
| Live Production | 120 requests | 60 seconds |
A retry is still an HTTP request and counts against the limit.
Headers
Authenticated responses include:
| Header | Meaning |
|---|---|
RateLimit-Limit | Maximum requests in the current window |
RateLimit-Remaining | Requests remaining |
RateLimit-Reset | Seconds until the window resets |
Retry-After | Seconds to wait after a 429 response |
Rate-limit response
{
"success": false,
"error": {
"code": "RATE_LIMIT_EXCEEDED",
"message": "Too many requests. Retry after the current rate limit window resets.",
"requestId": "550e8400-e29b-41d4-a716-446655440005"
},
"billing": {
"amount": 100,
"currency": "NGN",
"unit": "naira",
"charged": false,
"note": "No charges"
}
}Client behavior
- Read
Retry-Afterinstead of guessing a delay. - Add jitter when many workers share one key.
- Avoid retry loops for non-transient errors.
- Use separate API keys for independent applications when operational isolation is required.
Updated about 3 hours ago
Did this page help you?