ManagementQuotas
Handle quota exceeded
Understand and respond to 429 quota_exceeded.
Handle quota exceeded
When any active matching quota lacks capacity, the gateway rejects the request before the upstream provider call with 429 quota_exceeded.
What The Application Should Do
Treat quota exhaustion as a limit response.
Good responses include:
- slow down,
- wait for the next quota window,
- stop the current batch,
- ask the owner to raise the quota,
- investigate whether a deployment is sending unexpected traffic.
Avoid tight retry loops. The request will usually fail again until capacity changes or the window resets.
Troubleshooting Checklist
Identify the API key used by the failed request.
Check whether the key is organisation-, team-, or user-scoped.
Open Quotas and filter for matching organisation, team, user, and API-key quotas.
Open each active matching quota.
Review the current window's used plus reserved value.
Compare the consumed value with the configured limit.
Review Usage Records for the quota window.