Rate Limits

Quotas depend on your plan, model route and commercial agreement. Numerical RPM, TPM and concurrency limits are not published in this project.

Handling 429 responses

Respect Retry-After when provided. Queue work, cap concurrency and apply bounded exponential backoff with jitter. Avoid synchronized retry bursts.

Growth offers higher quota and priority routing. Enterprise customers can request dedicated capacity. Confirm the actual limits before load testing.