Rate Limits
Public API traffic is rate limited to protect platform availability and prevent abusive request patterns.
Current public limits:
| Layer | Rule |
|---|---|
| Request rate | Up to 5 requests per second per client |
| Burst behavior | Short bursts of up to 10 requests may be accepted |
| Connection guard | Excess concurrent connections may be rejected |
When a limit is exceeded, the API can return 429 Too Many Requests. Clients should treat 429 as retryable and back off before sending more requests.
Headers
No public rate-limit response headers, remaining-quota headers, reset timestamps, or endpoint-specific quota metadata are currently published by the API.
Client Behavior
- Keep request concurrency bounded.
- Retry
429responses with exponential backoff. - Do not retry unsafe write requests blindly unless your integration uses idempotency on its side.
- Avoid burst traffic during import, sync, or polling jobs.