I have a Google AI Ultra for Business account and I’m using the antigravity editor, and I’m noticing that the Google AI 3 Pro High model is taking a very long time to respond today. Just asking it what day it is takes quite a while, several minutes. I don’t know if this is happening to other people or if there’s some kind of issue. It’s impossible to work like this, especially since I have this type of account.
I’m having errors too =>
Error: HTTP 500 Internal Server Error
Sherlog:
TraceID: 0x898817ccafd7…
Headers: {“Alt-Svc”:[“h3=“:443”; ma=2592000,h3-29=“:443”; ma=2592000”],“Content-Length”:[“109”],“Content-Type”:[“text/event-stream”],“Date”:[“Wed, 18 Feb 2026 08:52:49 GMT”],“Server”:[“ESF”],“Server-Timing”:[“gfet4t7; dur=4758”],“Vary”:[“Origin”,“X-Origin”,“Referer”],“X-Cloudaicompanion-Trace-Id”:["898817ccafd74c21 "],“X-Content-Type-Options”:[“nosniff”],“X-Frame-Options”:[“SAMEORIGIN”],“X-Xss-Protection”:[“0”]}
Thank you for flagging this. Multi-minute response times on Google AI 3 Pro High aren’t expected, particularly on an Ultra for Business plan. This most likely points to temporary issues.
Could you please try to switch to the standard Pro model to see if the issue is isolated to Pro High. Also try opening a new session or launch the editor in an Incognito/Private window.
If performance doesn’t improve, please share your browser version, region, and when the issue began so we can escalate accordingly.
Both the high and low versions of the pro model are having the same problem. It works fine with Gemini 3 Flash. My location is Barcelona, Spain. It worked correctly yesterday, but today at 8:30 I started working and the pro model isn’t responding correctly from the Antigravity editor. I’ve tried logging out, disconnecting the account from Antigravity, and logging back in, but it’s still the same. The account I’m using isn’t the same one I use for the forum; it’s a work account with a Google AI Ultra for Business extension. And I keep getting the following error:
Error: HTTP 500 Internal Server Error
Sherlog:
TraceID: 0x32d1a8a2e2d…
Headers: {“Alt-Svc”:[“h3=“:443”; ma=2592000,h3-29=“:443”; ma=2592000”],“Content-Length”:[“109”],“Content-Type”:[“text/event-stream”],“Date”:[“Wed, 18 Feb 2026 09:08:52 GMT”],“Server”:[“ESF”],“Server-Timing”:[“gfet4t7; dur=371956”],“Vary”:[“Origin”,“X-Origin”,“Referer”],“X-Cloudaicompanion-Trace-Id”:[“32d1a8a2e2dc777a”],“X-Content-Type-Options”:[“nosniff”],“X-Frame-Options”:[“SAMEORIGIN”],“X-Xss-Protection”:[“0”]}
I noticed poor performance with Gemini Pro (High) all day yesterday: slow, inability to comply with direction, and poor decisions. I effectively lost the entire day wrestling with it. I’m on the Pro plan, in the US.
Same here but on personal Google One Premium plan.
Coincidence or not it seemed to die on the same day I got an email saying “Your Google One Premium 2 TB plan has been updated to include the same AI benefits as the [Google AI Plus] plan at no additional cost.” ?
This morning I saw Gemini Pro 3.1 and tested it. It was working fine for a while, but after a while it stopped responding at all; it just gets stuck (loading…, working…) but does nothing. I don’t know if anyone else is experiencing this. I have the new version and I’ve logged out, removed the application’s access from the account, and logged back in. I don’t know if anyone else is having this problem or if there’s something wrong with my account, but with this account, it should work fine. Our Plan: Google AI Ultra for Business
You are subscribed to the best Workspace plan.
I am having the exact same problem, it did not work fine for me at all.. for me its just stuck in a loop of “working… loading… “, most times I dont get a reply at all and when I do it can take upwards of 10 minutes even for a simple task using the “fast” mode
I am having the exact same issue since about a week back when Claude 4.6 was released. I am also on Ultra, and it doesn’t matter which model I use, Antigravity is extremely slow and hangs a lot, and has gone from a really good tool to being frankly unusable at the moment. Agents seem to also be performing worse, failing to close stale sessions and clogging system memory constantly. Never used to be this way.
I’m also experiencing the same issue. It’s painfully slow. It seems the actual thinking is fast, perhaps 1-4s, but then it gets stuck between thinking moments for minutes and therefore it looks like it’s fast, but is actually really slow!
I’m in the UK, and it’s been like this for a few days.
Flash works fine, but 3.1 pro high and low are extremely slow.
I just cancelled Codex PRO, and move to GEMINI with ANTIGRAVITY, first day was good, second day it stopped worked. Even I asked a ‘Hi‘, still get 500.
Have tried, quit antigravity and reopen, logout and re-log in, and nothing changed. Only flash can be used, but flash is model has its own idea. So I come back to re-subcribe a PLUS plan in Openai and wait ti see what will that problems go.
I mean, if you do not have enough place for users, you should stop let more users in and do not make payment and then after 1 day froze everything.
Can anyone tell me is that legal in US ?
Same here but i have noticed in my region if you use at the mid night it working better then the day time. i believe low user. but still it take 5 to 10 min in the morning it totally useless. like one thing 500 error. but right it 11pm in my country it working but slow.
I’ve been using google ultra since release, gemini 3.1 is an absolute disaster. this is the worst ai i have used in years. i am not sure why it is SO bad, but it is unbelievably slow, context rot is insane, it’s code is bad… what happened? I full on cancelled my ultra account and am using claude for everything atm. the claude (even sonnet) responses are still excellent and infinitely better than 3.1 (since 3.1 is completely useless..). very frustrating as i have month left on this useless service.