Features

Token-based rate limiting

This feature will be available in a future Kinference release.

Kinference allows you to limit tenant inference usage by input and output token rates.