503 UNAVAILABLE. {‘error’: {‘code’: 503, ‘message’: ‘The model is overloaded. Please try again later.’, ‘status’: ‘UNAVAILABLE’}}
I’m almost losing my mind over this, Not even a single request is passing by without having this shoved in my Face, Please fix your stuff, google
Hi @Vine,
I’m sorry to hear you’re running into this issue. This error means the model is temporarily overloaded due to very high traffic.
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.
Thank you!
Still hiting 80% of 503 errors for each API calls. How this is not a top priority to fix? I don’t understand. The problem has been there for 2 weeks. All my apps in production are unusable
been having the same issues for the last 2-3 weeks. 2.5-pro is basically useless; 2.5-flash is now reaching the same point.
Yes im using flash 2.5 and its now unusable. Crazy how noone gives a . They are still promoting their new models on X and telling how great they are while our apps are breaking since 2 weeks. Insane
I am having very much the same situation. It have forced me to pivot to other models.
Keep in mind that you’re dealing with a small startup level AI company and as such there might not be enough resources.
Anyway: having the same problem and seeing this kind of “solutions” here make me question our decision to use Gemini. The best quality doesn’t help if the servers are not equipped for everyday usage.
I am having the same issue with any models, I thought this was because I am using the free credit from google but it’s not, because I already setup my billing to pay as you go, this really the service problem.
same here why don’t they fix it fast? I’ve already check rate limit for tier 1 it’s far from it