Excessive 429 errors today – rate limit or availability issue?

Gemini API
1

Today we experienced an unusually high number of 429 errors when calling the Gemini API. As far as I understand, this error can be related to rate limits, quota exhaustion, or service unavailability.

A few questions:

  1. Was this a broader issue affecting other users today, or is it likely isolated to our account/project?

  2. When I check our quota under APIs & Services > Gemini API in the Google Cloud Console, everything appears to be within the established limits. Is that the correct place to monitor actual usage vs. availability?

  3. Is there a status page or recommended dashboard to track Gemini API availability in real time?

Any guidance would be appreciated. Thanks!

2

Hi givagov,

That is unexpected.

If you’re receiving 429’s and have checked no rate limit was breached for the period, please add your details to this form & we’ll check it out.

3

Hi Jon,

Thanks for the quick reply. Just to clarify — I’m actually calling Gemini through Vertex AI (genai.Client(vertexai=True, ...)), not through the Gemini API directly. Since the form explicitly states it’s only for Gemini API (not Vertex), I don’t think it applies to my case.

Could you point me to the right channel for reporting unexpected 429s on Vertex AI? I’ve already confirmed no quota limits were breached for the period, and I’m using region rotation across multiple Vertex regions plus the global endpoint for gemini-3-* models.

Happy to share request IDs, timestamps, and the specific regions/models affected if that helps.

Thanks!

4

If you have an assigned Field Sales Manager, they’d be the person to talk to. Otherwise the Welcome to the Google Cloud Community would be the place to go.