Today I got this new error for the first time. I have NEVER gotten this error before and I am just wondering if this is due to new extremely low quotas being implemented for the build apps with Gemini tool or if it’s just a bug. I have used the build mode frequently to test out various functionality since May, but now I can no longer use it due to the error appearing after just a few requests. Afaik there are no limits on the build with Gemini tool and I cannot find anything about it in the documentation. Neither is there any mention of what model the Build mode use nor a way for me to change model. The error is as follows:
“Failed to run code assistant, quota exceeded: you have reached the daily limit of requests for this model. Please select a different model, or try again tomorrow.”
Hey @Johannes_Wahlqwist,
Welcome to Forum!
that message means you’ve hit the daily usage limit for that Gemini model in Build mode. You can either switch to a different model or wait until your quota resets tomorrow. If this happens often, you might consider requesting a quota increase.
Thanks!
Hey, thank you for the quick reply. But did you read my post in its entirety? HOW do I change model in the build apps with Gemini tool? I know what the error means, but there is no way to change model or request a limit increase for the build apps with gemini tool, well, that I know of anyway. The build apps tool is a self contained tool with a default no named code assistant. Also when and why did google do this? If you can only do a few requests with the build apps with gemini, then how are users supposed to see what gemini can do?
Hey @Johannes_Wahlqwist , great follow-up! You’re right, the Build tool doesn’t currently offer a clear way to switch models or request quota increases directly. It’s a bit limiting, especially for exploring Gemini’s full potential. Hopefully Google adds more flexibility soon so users can really test things out without hitting walls so quickly. Thanks for pointing this out!
But why hitting users with quotas now all of a sudden without any documentation or ANY form of mention of this what so ever?? I have used this tool for months heavily without ever hitting any quotas or anything.
Hey @Johannes_Wahlqwist , that’s a totally fair point. If quotas were introduced recently, it really should’ve come with clear communication or at least a heads-up in the docs or UI. Sudden changes like this can be super frustrating, especially for regular users like you who rely on the tool. Hopefully the team clarifies this soon and adds more transparency around usage limits. You’re absolutely right to raise this.
Thanks for speaking up!
I actually have the same problem! “Failed to run code assistant, quota exceeded: you have reached the daily limit of requests for this model. Please select a different model, or try again tomorrow.”
But no option to change the quota or change the API key or whatever. And the problem is that I use AI Studio to create a React App, yet this has stopped because I keep getting this error. And I got it yesterday too. My quota seems very small, while I know that I have set a big amount to use for the API services.
As my project is racing towards a deadline, I now have additional problems making this work. Suddenly, this great tool has become unusable.
And what is weird is that some of my projects also use the Gemini API, yet they don’t hit any quota! I have four keys that can be used and each should have a big quota. So it seems that AI Studio is using some free API key instead, and thus a limited quota. I need it to use my real key instead.
Google, take my money as this service is valuable to me! Don’t give me a limited freebee. 
Hello,
Welcome to the Forum!!
Could you please share on average how many requests you process before getting this error and are you working with an existing application or new one?
Give Dyad a look—you can import an existing app with almost no friction. I originally favored Gemini Build Apps because deployment was effortless, but Dyad is quickly closing that gap. My billing is already configured, yet the platform still doesn’t warn me when I hit the free-tier limit or prompt me to upgrade (Flash vs Pro); hopefully that UX tweak is on the roadmap. For many projects Firebase is overkill—we need an in-between option. Dyad’s ability to write directly to Supabase strikes a good balance and unlocks some excellent features.
I noticed that I myself run out of quota/resources in my app when my App uses gemini-2.5-flash-preview-04-17 to call Gemini. But when I tell AI Studio to use gemini-2.5-pro instead, I have no problems any more.
I am working on various projects at the moment, where AI Studio has to create apps that call Gemini again for various processing. Often related to coding tasks, as I use AI Studio to make simple apps to solve specific coding problems, like building data models with data seeding where the App creates the C# DBContext and fills it.
Or Apps that have to analyze images.
It’s weird that Flash quickly runs out while Pro lasts.
Another thing I’ve noticed is that I created a React App in AI Studio and inside the sandbox, the app gets in trouble with the resources. Ir runs out of them. But I deployed the same app to my own server, and there I have no problems. Plenty of quota there, apparently.
This app calls Gemini 2.5 Pro and on start-up, I have to provide the API key, which it will store in local storage. I use the same key in the app AI Studio and on my server. (Keeping the key in my browser only, where the app is running.) So it should have the same restrictions inside AI Studio as outside.
However, outside I can keep using the AI while AI Studio starts blocking me, telling me I ran out of resources, quota or whatever.
So there seems to be a limit in AI Studio regardless the engine used.
Your quota depends on the number of token and number of request you make, for Gemini API you can find this information here.
When does the 24-hour quota period reset?
At 4:00 PM AEST I commenced receiving “Failed to generate content, quota exceeded: you have reached the daily limit of requests for this model. Please select a different model, or try again tomorrow.” Followed my many fail to generate errors. I’m assuming I had reached the 2.5 Pro 100 RPD over 2 consecutive chat threads. The thread the error occurred in I was approx. 88 requests.
I had never received this message before in my many sessions, so I’m attempting to confirm if this is a Rate Limit error or the same issues that I have been experiencing this past month.
Thank you
Yet there’s a difference in the quota while using AI Studio and when just using my own deployed site. Same API key, same project and everything. My app asks the user to provide this API key and then uses this specific key. But in AI Studio, it still tells me I run out of resources. (Status code 429) while my site hasn’t encountered it.
As I am developing this app at the moment, I use them both at the same time for testing purposes. But once AI Studio gets the 429, I have to stop development as testing is now failing. That’s annoying!
For my sessions with 2.5 pro between 100 - 200k they commence becoming unstable for me now, so I keep starting new threads but it is such a drag on any development work, I havent deployed any yet as the system interruptions are certainly halting progress. thanks for responding.