Rate limited

Status

429

Retryable

yes

Type

https://fastpad.ai/errors/rate-limited

What it means

Too many requests in the window, or more than this server can take at once. Depending on the route, the limit is per token, per person or per machine — for example a token’s own allowance, the `/v1/ask` allowance a person’s session and every CLI token they hold draw on together, or one server’s budget for large request bodies. Whichever answered, this is a throttle, not the service being down.

What to do

Wait `retry_after_ms` from the response body, then retry. That value is what to honor whichever limit answered; a retry sent sooner is refused again.