10 RPM quota being applied to Paid tier (gemini-2.0-flash-lite-preview-02-05)

I read from 429 Resource has been exhausted (e.g. check quota) for Paid Gemini Flash - #2 by OrangiaNebula that the quota should gradually increase as monthly payments are made. But the bigger issue is that the 10 RPM is not documented (which makes paid tier 1 worse than the free tier) and feels so low that it’s probably a bug that is blocking prod usage of gemini.

Update: My dev key is paid tier 2 (4M RPM) and my prod / staging keys are paid tier 1 (the 10 RPM kind).

This 10 RPM for paid tier 1 means we cannot easily cycle new keys in (e.g. if a key is compromised)

2 Likes

The gemini-2.0-flash-lite requests are now going to the proper paid tier 1 quota (4000 RPM) :+1:

Facing same issue.

Update: gemini-2.0-flash-lite-preview-2 is hitting 429 (Rate limit) around ~10 RPM but now the usage isn’t showing in the quota page (https://console.cloud.google.com/apis/api/generativelanguage.googleapis.com/quotas).

Same problem here

Getting 429 error with gemini-2.0-flash and gemini-1.5-pro. I am on paid tier and making less than 100 RPM

very frustrating

gemini-2.0-flash-lite-preview seems to map to the correct paid tier quota “gemini-2.0-flash-lite” so the queries are now going through as expected