Introduction

If you're working with the Google Gemini API on the free tier and suddenly start seeing a System.InvalidOperationException combined with an HTTP 429 (Quota Exceeded) error, you are not alone. This is one of the most common issues developers face when getting started with Gemini AI — especially when using models like gemini-3-pro-image or other generative language models on the free plan.

In this blog post, we'll walk through exactly why this error occurs, what the free tier limits look like, and how you can fix it quickly. We also have a full video walkthrough below — check it out!

Watch the Video Tutorial

What is Gemini API Error 429?

Error 429 (RESOURCE_EXHAUSTED) means you have exceeded the quota or rate limits set for your Gemini API account. This happens because the Google Gemini free tier has strict limits on how many requests you can make per minute, per day, and per model. When your app exceeds these limits, the API returns a 429 status and throws a System.InvalidOperationException in .NET or similar exceptions in other languages.

Understanding Free Tier Rate Limits

The Gemini free tier is generous for experimentation but has hard ceilings. Key limits include:

• Requests Per Minute (RPM): Limited to a small number of requests per model per minute.
• Tokens Per Minute (TPM): Total input + output tokens are capped per minute.
• Requests Per Day (RPD): There is a daily cap on the number of API calls.
• Image Generation Models: These models have even stricter limits and may hit 0 quota on the free tier very quickly.

You can monitor your current usage at: https://ai.dev/rate-limit

How to Fix Gemini API Error 429

1. Implement Retry Logic with Exponential Backoff

The best immediate fix is to implement a retry mechanism with exponential backoff. Instead of letting your app crash on a 429, catch the error and wait before retrying:

int maxRetries = 5;
int retryCount = 0;
TimeSpan delay = TimeSpan.FromSeconds(2);
while (retryCount < maxRetries)
{
    try
    {
        var response = await geminiClient.GenerateContentAsync(request);
        break;
    }
    catch (Exception ex) when (ex.Message.Contains("429"))
    {
        retryCount++;
        if (retryCount == maxRetries) throw;
        await Task.Delay(delay);
        delay = TimeSpan.FromSeconds(delay.TotalSeconds * 2);
    }
}

2. Optimize Your API Calls

• Batch requests where possible instead of making many small calls.
• Cache API responses for repeated or similar queries.
• Reduce the frequency of API calls in your application logic.

3. Monitor Your Quota Usage

Regularly check your API usage dashboard at ai.dev/rate-limit to understand how close you are to your limits.

4. Upgrade to a Paid Plan

If your application requires consistent, high-volume API access, the free tier may not be sufficient. Consider upgrading to a paid Gemini API plan which offers significantly higher rate limits. This is especially important if you are using image generation models like gemini-3-pro-image.

5. Using Image Models on the Free Tier

The Gemini image generation models have very low or even zero quota on the free tier. If you are seeing this error specifically when trying to generate images, you will need to either upgrade to a paid plan or reduce the frequency of image generation requests drastically.

Video Timestamps

• 0:00 - Introduction
• 0:30 - What is Error 429?
• 1:30 - Free Tier Quota Limits Explained
• 3:00 - How to Fix It (Code Walkthrough)
• 5:00 - Best Practices to Avoid This Error

Best Practices to Avoid This Error

• Always implement retry logic with exponential backoff in production apps.
• Set up quota alerts in your Google Cloud console.
• Use caching aggressively to reduce redundant API calls.
• Design your application to gracefully degrade when API limits are hit.
• Consider a paid tier if your use case requires reliable, consistent API access.

Conclusion

The Gemini API Error 429 with System.InvalidOperationException is a common hurdle on the free tier, but it is entirely fixable. By understanding your rate limits, implementing proper retry logic with exponential backoff, and optimizing how your application makes API calls, you can build resilient AI-powered applications even on the free plan.

Watch the full video tutorial above for a live code walkthrough and demo of these fixes in action. Don't forget to like and subscribe to Code2Night for more .NET and AI tutorials!