> ## Documentation Index
> Fetch the complete documentation index at: https://docs.layarva.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Queue API Rate Limits

> Design Queue API and device clients around request, nonce, and PIN limits.

## Queue API

* Default shared credential limit: 60 requests per one-minute bucket.
* A rejected over-limit request returns HTTP 429 with `RATE_LIMITED`.
* Use exponential backoff with jitter and wait at least until the next minute bucket.
* Do not retry a ticket mutation with a new Idempotency Key after an unknown network result; retry the same logical operation with the same key.

## Replay records

Mutation nonces are retained for ten minutes. Request timestamps must be within five minutes.

## Device PINs

Screen unlock, Screen release, and Kiosk unlock use five attempts per five minutes per device token. Successful verification clears the counter.

## Related guides

* [Authenticate Queue API Requests](/developers/authentication)
* [Secure and Release a Screen](/screens/device-security)