Issues with Gemini API: Rate Limit Errors and Request Surges

Ksenia B. · 2025-12-09T16:22:29.511Z

I have been having issues with Gemini for almost 2 weeks now, I'm on a paid tier and currently trying to use it only in 1 table (https://app.clay.com/workspaces/261968/workbooks/wb_0szqs5hGwRDaFotGWiV/tables/t_0t6n24bhEQGGMDEuw5t/views/gv_0t6n24cxbfJSchex5CC), outside of Clay I have 100 requests per day max. Every time I try to run even one row in an hour, I'm getting an error: "The model provider's rate limits are low. Please try again later or increase your limits." even though we're not hitting any model limits (Current limit is 2k requests per minute). I also see in the Gemini dashboard that Clay is sending way more API requests than we trigger with 50%+ resulting in an error. Can someone please help us fix it? More than a week ago everything was working just fine.

Issues with Gemini API: Rate Limit Errors and Request Surges | Clay

11 comments

· Sorted by Oldest

Clay Support
APP
·
·
This sounds like a Gemini API tier issue. For Gemini to work reliably in Clay, you need Tier 2 access, and it only works with the Gemini 2.0 Flash model - older Gemini 1.5 models may require Vertex AI or custom tiers. [1]
The extra API requests you're seeing are likely Clay's retry mechanism and smarter pacing features that help handle rate limits better. With lower API tiers, you'll need to process smaller batches - try running just a few rows at a time instead of larger batches.
Check your Gemini API tier at the Gemini support page and consider upgrading to Tier 2 if needed.
Before we escalate this, have you tried using Sculptor (top-right corner) to help troubleshoot your workflow? It's free and might help identify the specific configuration issue.
Ksenia B.
·
·
Not helpful and you don't event have Gemini 1.5
Clay Support
APP
·
·
I'm sorry to hear that you're frustrated. I'll connect you with a human agent who can assist you further.
Clay Support
APP
·
·
Our support team has got your message and we'll get back to you soon!
If you’re dealing with a specific table, drop the URL below so we can help you quicker. Otherwise, someone from our team will be in touch soon!
image.png(93 kB)
Ksenia B.
·
·
Is it possible to turn off Clay retry mechanism?
Andrew Q.
·
·
Hey Ksenia! Thanks for reaching out. Could you clarify what you're referring to? Do you mean auto-update?
Ksenia B.
·
·
I was referring to the support answer the bot sent above. I'm seeing thousands of API request calls to Gemini while I trigger 10x less in Clay, how come?
Ksenia B.
·
·
Any updates Andrew Q.? I'm seeing thousands of requests on Gemini end while triggering very few rows in Clay. One row is stuck for 7+ min probably being retried on the back for 100+ times
Andrew Q.
·
·
Hey there! Gemini is something we're working on right now (especially for larger rows) but we have mitigated the problem for our key. Since this is a personal key that you're using, have you reached out to Google to upgrade keys? At the moment, have you considered running a different model if you have other keys? I recommend Anthropic or an OpenAI key if you have one.
Alex B.
·
·
Andrew Q. was this resolved? I am having the same issue.
Kashif J.
·
·
Hi Ksenia, Our engineering team is currently working on this and we'll be sure to get back to you once we have an update. Thank you for your patience

Clay Support
APP
·
·
This sounds like a Gemini API tier issue. For Gemini to work reliably in Clay, you need Tier 2 access, and it only works with the Gemini 2.0 Flash model - older Gemini 1.5 models may require Vertex AI or custom tiers. [1]
The extra API requests you're seeing are likely Clay's retry mechanism and smarter pacing features that help handle rate limits better. With lower API tiers, you'll need to process smaller batches - try running just a few rows at a time instead of larger batches.
Check your Gemini API tier at the Gemini support page and consider upgrading to Tier 2 if needed.
Before we escalate this, have you tried using Sculptor (top-right corner) to help troubleshoot your workflow? It's free and might help identify the specific configuration issue.
Ksenia B.
·
·
Not helpful and you don't event have Gemini 1.5
Clay Support
APP
·
·
I'm sorry to hear that you're frustrated. I'll connect you with a human agent who can assist you further.
Clay Support
APP
·
·
Our support team has got your message and we'll get back to you soon!
If you’re dealing with a specific table, drop the URL below so we can help you quicker. Otherwise, someone from our team will be in touch soon!
image.png(93 kB)
Ksenia B.
·
·
Is it possible to turn off Clay retry mechanism?
Andrew Q.
·
·
Hey Ksenia! Thanks for reaching out. Could you clarify what you're referring to? Do you mean auto-update?
Ksenia B.
·
·
I was referring to the support answer the bot sent above. I'm seeing thousands of API request calls to Gemini while I trigger 10x less in Clay, how come?
Ksenia B.
·
·
Any updates Andrew Q.? I'm seeing thousands of requests on Gemini end while triggering very few rows in Clay. One row is stuck for 7+ min probably being retried on the back for 100+ times
Andrew Q.
·
·
Hey there! Gemini is something we're working on right now (especially for larger rows) but we have mitigated the problem for our key. Since this is a personal key that you're using, have you reached out to Google to upgrade keys? At the moment, have you considered running a different model if you have other keys? I recommend Anthropic or an OpenAI key if you have one.
Alex B.
·
·
Andrew Q. was this resolved? I am having the same issue.
Kashif J.
·
·
Hi Ksenia, Our engineering team is currently working on this and we'll be sure to get back to you once we have an update. Thank you for your patience