Frequent Gemini 2.0 API errors - 429, 503 (parallel processing)

Hi, I’m experiencing frequent rate limit errors while using the ‘gemini-2.0-flash-001’ model for text classification. I have multiple API keys from separate Google Cloud projects and am using a parallel pool executor to speed things up.

Despite this setup, I’m consistently receiving errors like the ones shown here [429 and 503]:

However, as you can see from my API usage metrics below

My API usage is nowhere near the per-minute rate limit (or even the per-day limit) so I’m very confused.

Furthermore, I’d appreciate some clarification on ‘v1beta’ in the error message: ‘google.ai.generativelanguage.v1beta.GenerativeService.GenerateContent’.

  • What does ‘v1beta’ signify in this context?
  • Is it possible to switch to a different API version for this specific method while still using the ‘gemini-2.0-flash-001’ model?

Any insights or suggestions would be greatly appreciated. Thanks!

3 Likes

This sounds related to the issue I started experiencing last week as well:

1 Like