As of March 2, 2026, there have been numerous reports from the developer community regarding “Internal Error” messages, hanging requests, and session persistence failures.
This is largely due to the ongoing transition and backend updates related to Gemini 3.1 Pro.
Current Status & Known Issues
Platform Stability: Users are reporting that sessions frequently crash or return “Internal Error” after a few turns of conversation.
“Infinite Thinking”: The model may get stuck in a “Working” state for several minutes before failing.
Recent Outages: Official status logs show intermittent issues with the AI Studio “Build” environment and the Gemini API throughout late February and into today.
Troubleshooting Steps
If the error is persistent, you can try these community-verified workarounds to bypass the current backend glitches:
Reset the Session:
Many internal errors are tied to “corrupted” session states. Try creating a New Prompt or New Chat rather than continuing an existing one.
Model Switching:
If you are using
gemini-3.1-pro-preview, try switching to the more stablegemini-3-flashorgemini-1.5-proto see if the issue persists.
Clear Browser Data:
Open AI Studio in an Incognito/Private window. If it works there, you likely need to clear your browser cache and cookies for
aistudio.google.com.
Check for “Creative” Conflicts:
Some users have found that high Temperature or Top-P settings trigger internal errors during the preview phase. Try lowering Temperature to 0 and Top-P to 0.9 to see if it stabilizes the response.
File Size Limits:
Ensure any uploaded images are under 7 MB and JSON/PDF files are relatively small. Large attachments are currently a known trigger for the “Internal Error” bug.
Also Read : Unknown Error Occurred ChatGPT File Upload
Official Status Links
You can monitor the situation directly through these official channels:
Since this is a backend issue on Google’s side, it often requires waiting for a few hours while their engineers roll out a patch.
Be the first to comment