Incorrect tier's rate limits being applied?

Gemini API
,
1

Hi,

We are using gemini-2.5-flash (and 2.5 pro and 2.5 flash-lite) with a tier 3 api key (from https://aistudio.google.com/u/1/apikey). However we occasionally hit rate limits earlier than expected with the following response message:

ClientError: 429 RESOURCE_EXHAUSTED. {'error': {'code': 429, 'message': 'You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits.', 'status': 'RESOURCE_EXHAUSTED', 'details': [{'@type': 'type.googleapis.com/google.rpc.QuotaFailure', 'violations': [{'quotaMetric': 'generativelanguage.googleapis.com/generate_content_paid_tier_2_input_token_count', 'quotaId': 'GenerateContentPaidTierInputTokensPerModelPerMinute-PaidTier2', 'quotaDimensions': {'location': 'global', 'model': 'gemini-2.5-flash'}, 'quotaValue': '3000000'}]}, {'@type': 'type.googleapis.com/google.rpc.Help', 'links': [{'description': 'Learn more about Gemini API quotas', 'url': 'https://ai.google.dev/gemini-api/docs/rate-limits'}]}, {'@type': 'type.googleapis.com/google.rpc.RetryInfo', 'retryDelay': '19s'}]}}

I’m confused why this message mentions PaidTier2 instead of tier 3, and the TPM is 3M instead of 8M. Is the mention of rate limits a red herring here, and this is just a typical 429 resource exhausted error? Or are we incorrectly having the tier 2 TPM limit applied?

1 Like
3

Hi @ascillitoe,

Welcome to the Forum,

If your Tier 3 API key is restricted to tier 2 rate limits, please contact the dedicated Billing Support team for quick assistance. They can be reached directly via this link Get Cloud Billing support  |  Support Documentation  |  Google Cloud.

Thank you!

1 Like
5

Thanks for the welcome @chunduriv !

The billing team’s response was:

I understand your confusion about why the error appears every time your usage reaches 3M, despite the rate limit for Tier 3 is 8M. Regarding this, please be advised that further investigation about the error will require technical expertise, and since I’m from Google Cloud Billing, I have limited knowledge of such errors. I can only provide basic information. With this, I will need to seek assistance from our Technical Team for further investigation. I kindly ask for your understanding on this matter.

So not particularly helpful…

7

Just to follow up @chunduriv, technical support said they could only advise on Vertex AI issues and I should post at discuss.ai.google.dev. So a big waste of time!