429 · Rate limited
rate_limit_exceeded normally identifies the gateway's per-key limit; upstream_rate_limited identifies an upstream limit. Neither proves model retirement.
On this page
What this means
rate_limit_exceeded normally identifies the gateway's per-key limit; upstream_rate_limited identifies an upstream limit. Neither proves model retirement.
429 · rate_limit_exceeded / upstream_rate_limitedCheck and resolve
Honor Retry-After when present; upstreams may omit it. Otherwise use bounded exponential backoff with jitter. Reduce concurrency or frequency and avoid immediate repeated retries or rotating keys to evade limits.