Unreliable API requests for Gemini 2.5 pro and Gemini 3.0 preview

I work via a Google AI Studio API via n8n to generate an image using gemini-3 pro image preview (Nano Banana Pro). Back-up model is gemini 2.5 flash image (Nano Banana). However both give me errors. It is very unreliable. How can I fix this?

Error:

{ “error”: { “code”: 503, “message”: “The model is overloaded. Please try again later.”, “status”: “UNAVAILABLE” } }

Hi @surfer , Thanks for reaching out to us.

A 503 error (UNAVAILABLE) indicates that the service is temporarily overloaded or experiencing a capacity constraint. This is common during peak usage hours, particularly with long context requests.

The best immediate step is to wait a few minutes and try again . If you’re using code, implementing a retry with exponential backoff is the most robust way to handle this.

This seems to be the generic answer to everyone. It’s been days that it returns the same error no matter the time of the day

I’m experiencing the same issue in my web app that uses the Gemini API with NanoBanana Pro.

I’d like to know whether Google is aware of this issue and plans to address it, or if this is expected behavior due to the high demand for these models.

If the instability is demand-related, have you identified any specific time windows when the service is more stable or performs better? This would help me organize my workflow and coordinate with my team. @Sonali_Kumari1

Hi everyone,

could you please let us know if you are using the free tier or a paid tier? This information is helpful because requests from paid tiers are prioritized, especially during periods of high traffic.

I am using paid tier 3, and have this 503 problem for the whole day by using batch API & file API for processing video understanding.