I’m fed up with a problem that’s ruining my experience using Antigravity. Maybe someone else has encountered something similar. I have no idea what exactly might be causing this issue.
I can’t keep chats going consistently because sometimes I run into a server-side freeze. Here are the symptoms I’m seeing:
- The “Generating” log flashing
- Quotas and limits get used up if I don’t stop the generation within 5–10 seconds (This exhausts the weekly quota)
- It’s impossible to continue the chat (there will always be a loop with the “Generating” log flashing)
I had a hunch that this might be related to large chats with a lot of text (though that’s nonsense—the model’s context window handles that volume of tokens without any issues). However, I encounter this problem in chats regardless of their size. For example, large chats have crashed there several times, and a couple of times—chats that had barely started.
I don’t perform any specific actions. Basically, I just have a chat going, text generation, and conversation; periodically, I ask the model to read a couple of small files or make changes.
Has anyone else encountered this problem? Do you know a solution? Maybe there’s a thread I missed?
So far, I’ve only encountered this issue on Windows 11. It doesn’t depend on the version of Antigravity itself. The only clue is that in February, everything was completely stable; the problems started sometime in March.
On one single occasion, the model broke out of this generation cycle and returned a response after a couple of minutes, but I was charged 500 AI credits and 40% of my Gemini 3.1 limit for that.