Gemini Block RPM limiting

Hello, I had a question regarding Gemini Block. Is there a way to limit requests per minute (RPM) when using a Gemini Block?

I’m currently running Gemini inference on each crop from Dynamic Crop with a free-tier API key that has a low RPM.

Is there a setting similar to max_concurrent_requests for controlling RPM, or is implementing this via a custom Python block the only option at the moment?

You could try the Rate Limiter Block (but note that this is local state so on Serverless v2 where there are many backends that may serve the request this will be per-server vs a global rate limit).

1 Like

Got it ill try this.