I think whole AI Studio is creating glitches especially Gemini 3.1 Pro, taking extra time and I feel 3.0 did better work last time, but recent roll-outs completely vanished all progresses. I think elsewhere 3.1 pro is working fine, but causing issues in AI studio only.
These are the problem’s I notices since last week & constantly from last 4 days:
Extreme Delays
Lags
First time, I encountered daily limit used notice. Even when I did not used that much compared to last days. I seems AI keeps thinking even after ending one process and results in draining of the usage.
Gemini’s understanding context ability weakened after this update
Data or Code losses. Lot of code was literally wiped out.
Thinking or structuring ability also weakened.
Before with small or moderate prompts, the results literally were perfect. But now long prompts or detailed contexts are not improving, just worsening it.
Even with correct prompts, It looks like it’s not performing the correct actions or not taking commands seriously.
After completing tasks, an error always appears which prompts to ‘fix’ always. But it does not fix anything and keeping things in loop.
Editor does not show active code writing page by page, which it used to do before.
I tried to provide a detailed context on this thread so Google’s Technical Team may find these useful since I’m using AI Studio since a long time and recently has been very active in analyzing these.
I hope Google will fix these very soon. Kindly feel free to tag Google’s team in this thread, this could be lot time-saving for them.
the only thing i keep doing is making a remix of my app so it has a fresh context window.
Every time i create a new implementation and publish the change i then do another remix and move on.
This is just a workaround that works for me, it should not behave like this.
Has nothing to do with the model i guess, happens for me with every single model possible. Loading years to a point where it stops with a error message. Or whats even better: Instantly giving me error messages. Today it was also WITH free qouta aswell as with paid key. Its only getting worse at this point and we have zero communication.
Is it working properly for everyone yet, last nights quota exceeding was strange since I had barely used the platform without it erroring out in some shape or form.
Interested in knowing how it’s acting for everyone else today?
Today every time i try to continue coding with Gemini 3.1 pro preview it successfully starts to work through the situation but always ends up returning an error after about 1300 seconds.
