# How should clients handle rate limits and retries?

> V2 applies rate and concurrency limits; respect HTTP 429 and Retry-After. Every repeated prediction is a new operation with normal billing. Do not automatically retry a timeout, lost response or unconfirmed billing outcome. Inspect code, action and meta.usage.billing_status first; contact support with request_id for billing_reconciliation_required. Use bounded backoff only when billing is confirmed and the documented action permits a new attempt.

Canonical HTML: https://www.genderapi.io/frequently-asked-questions/rate-limits-and-retries

Last reviewed: 2026-09-26

## Answer

V2 applies rate and concurrency limits; respect HTTP 429 and Retry-After. Every repeated prediction is a new operation with normal billing. Do not automatically retry a timeout, lost response or unconfirmed billing outcome. Inspect code, action and meta.usage.billing_status first; contact support with request_id for billing_reconciliation_required. Use bounded backoff only when billing is confirmed and the documented action permits a new attempt.

## Related canonical guidance

- [Errors, limits and retries](https://www.genderapi.io/docs/v2/errors-and-retries)
- [Credits and usage](https://www.genderapi.io/docs/v2/credits-and-usage)