Limited request per minute (5 RPM) for Gemini models

Google Cloud Build with AI AI APIs
1

I am using the latest Gemini models to run some research projects with videos and need to process multiple video frames in a sequential manner. But I find out both gemini-1.5-flash and gemini-1.5-pro only have a limited request per minute = 5.

In Google AI’s advertisement, they claim to have a 360 RPM limit…

Could you help me double check on the RPM limit? This is a serious roadblock to me

10 Likes
2

Facing he same issue. Usage doesn’t even show up in the quota

3

The rate limits for using the Gemini models depend on whether you are using the free tier or the paid tier. Here are the specifics:

  1. Free Tier:

    • Gemini 1.5 Pro: The free tier offers a limit of 2 requests per minute (RPM) and 50 requests per day (RPD)​.
    • Gemini 1.5 Flash: This model has a higher free-tier limit of 15 RPM and 1,500 RPD.

  2. Paid Tier:

    • Both Gemini 1.5 Pro and Gemini 1.5 Flash can handle up to 360 RPM and 10,000 RPD on the paid tier​.

If you’re experiencing limitations with the free tier, you might consider upgrading to the paid tier, which offers significantly higher request limits. This should help alleviate the roadblock in your research projects, especially if you’re processing video frames sequentially and need higher throughput.

2 Likes
4

How to switch to paid tier?

1 Like
5

On my Quotas & System Limits for project page, for Vertex AI API - Generate content requests per minute per project per base model per minute per region per base_model I have the value 1. How can it be? In the description for my region, it says 60 request per minute. I pay a billing enabled account and I pay for other services, Translation for example, every month. Please help me solve the problem.

6

What happens when I send 3 requests per minute and still get response, does gemini switch to lower models then?