I’m currently facing an issue with the Gemini API and hoping for some clarification from the team or anyone experiencing the same.
I keep receiving the following error message when making requests:
“Quota exceeded. Please try again later.”
I really don’t understand why this is happening because:
I am using a paid API key (Pay-as-you-go).
My billing is active and I am completely off the free tier.
My quotas in the Google Cloud Console are set to the maximum available.
I am aware that there have been ongoing issues over the last few days. Previously, I was consistently getting “internal error” messages, but now it seems to have shifted to this quota issue.
Did Google impose a new strict limit to handle the recent server instability? As it stands, it feels impossible to work reliably with paid keys right now.
Is anyone else experiencing this exact shift from internal errors to quota limits, and is there any official word on when we can actually use the capacity we are paying for?
I am also having the same issues on a project I have been working on for a good while.
Every since the migration I’m having issues with the API Paid Tier 1. The free version works but its answers and context are not anywhere as good for my current project.
It’s already been a week+ with the same issue. Hope they can resolve this soon.
Google AI Studio is the absolute “Modified by moderator” of the century.
Why launch a product you don’t understand, just to lure customers to you?
For several days now, nothing has been working, all those dream promises just to deliver this. And to top it all off, no communication, no information whatsoever for users—what a load of rubbish.
That doesn’t seem to work for me. I’ve read the Read Limits documentation amd checked the limits on my project’s Paid Tier, and the stats suggest I’m nowhere near the limits (see scrnshot), yet I still get “Quota exceeded. Please try again later.“
Quotas configured to the maximum allowed in Google Cloud Console
Despite this, I consistently receive the error:
“Quota exceeded. Please try again later.”
This started shortly after the migration. Previously, requests were failing with “internal error,” and now those have seemingly transitioned into quota errors.
From the outside, this does not appear to be a true quota issue. Usage is well within limits, and billing is active. The behavior suggests there may be backend throttling or temporary capacity restrictions being surfaced as quota errors.
This creates a serious reliability problem for production systems. If paid tier users cannot depend on stable API access, it becomes very difficult to build or maintain applications on top of the service.
