I am writing this with extreme frustration and disappointment regarding the Google AI Pro subscription service on the Antigravity platform.
We pay for “Pro” tiers to get reliability and access. Instead, I am facing what feels like a complete scam or a broken system that penalizes paid users without logic.
The Situation: I am a subscriber to the Google AI Pro (2 TB) plan. The explicitly stated policy for model usage (specifically for Claude Sonnet 4.5 / Opus 4.6 on Antigravity) is a reset cycle of 5 HOURS.
However, the reality is shocking:
The 167-Hour Lockout: As of today, my account shows a reset timer of 167 HOURS (approx. 7 Days). This is not a glitch; this is a full week’s ban.
False Quota Display: My status bar clearly shows that I have 60% Quota Remaining. I haven’t even used my limit! Yet, the system has locked me out for a week.
Recurring Nightmare: This is not a one-time error. Just a few days ago, I was locked out for 4-5 days arbitrarily. Now, immediately after getting access back, I am slapped with a 7-day lock again without heavy usage.
Evidence:(Please see the attached screenshot showing the 167h 52m reset timer on a Pro account)
My Question to the Google/Antigravity Team:
Is this a Shadow Ban? If so, why? I am a paying customer following the rules.
Is this a Hidden Policy Change? Have you silently changed the 5-hour reset to a Weekly Reset for high-end models without informing users?
Why pay for “Pro” if the service is worse than a free tier?
This level of instability makes the platform unusable for professional work. I demand an immediate explanation and a fix for my account.
To the Community: Has anyone else faced this ridiculous “167-hour” wall? Please comment below if you are experiencing similar issues. We need to bring this to the developers’ attention immediately.
same issue has been happening with me as well since so many weeks. I took another subscription from a different account thinking that my first account was faulty but i encountered the same issue in that as well by the 3rd day. This is a consistent issue that is persisting since so many days and many people have created so many support tickets yet no proper response. some google guy shows up and says that these limits only affect very very few set of people ( my entire freaking company faces the same issue!!! ) At this point this is a very unreliable tool for doing coding tasks. And more over if you use gemini models after getting 160hrs of limit on anthropic models they will also lock you out for another 160hrs leaving you with zero ai models which is extreme utrocity at this point. i am looking out for other alternativies, its better you also take this as a lesson and move forward. these guys are not even listening to us.
I’m facing similar issues, and it’s seriously hurting my daily work.
The Claude models actually get my tasks done and fix bugs quickly, but once the limits hit, I’m forced to use the Gemini models, and the quality drops a lot. Instead of fixing problems, they often create new ones and waste more time.
If paid users are getting locked out for days with no clear explanation, then the “Pro” plan doesn’t feel reliable at all. We just need stable limits and clear communication.
It also happened to me once ( I have AI Pro subscription for about 2 weeks ), but now I got banned from Gemini/Antigravity with no solution. After 5 days of contacting support and receiving some basic troubleshooting stuff, I decided to cancel my subscription. Somehow I am grateful because this happened in the first free month.
I am experiencing a bug with the quota tracking on my AI Pro subscription in Google Antigravity IDE.
Despite not using Gemini 3 Pro for the past 24+ days, my account is currently showing that I only have a 20% quota remaining, accompanied by a 6.5-hour reset window.
This calculation is entirely broken. After a day of zero activity, my quota should be fully restored. Instead, my tier limits are being throttled as if I have been actively using the system.
We should claim our payment back, because this is just not done. Is this a joke? Google should have some answers for us. I have reported this issue several times through Antigravity’s report system, but there has been no feedback, nothing. I guess they just don’t care about the problem their system is causing.
Also, the agent’s power to solve problems was great back then, but now it sucks. Sometimes I feel like, what the heck is it doing? I commanded something, and it is doing another thing, and also, it gets slower and slower.
I’m facing the exact same 167-hour lockout on a Pro account despite having quota left — this feels completely deceptive and unacceptable for a paid service, and the Antigravity team needs to explain this immediately.
CRITICAL ISSUE: Google AI Pro Plan - Unexpected Weekly Rate Limit (7-Day Lockout) instead of 5-Hour Reset
I am a paying subscriber to the Google AI Pro plan. Based on previous service behavior and documentation, my usage quota is supposed to refresh approximately every 5 hours.
However, since the recent update, the service has become unusable for professional work.
The Issue: After sending only 5 to 6 prompts using high-reasoning models (specifically Gemini 3 Pro and Claude Sonnet), I am immediately hit with a 48-hour to 7-day lockout (Weekly Rate Limit).
This happens even though the UI status bar indicates that I still have quota/tokens available.
Impact: I am effectively locked out of the service I pay for after less than 1 hour of work.
Please clarify if this is a technical bug regarding account flagging or an undocumented policy change. If this is intended, it renders the Pro plan unfit for development purposes.
I request that you investigate my account rate limits and restore the standard daily/5-hour refresh cycle immediately.
Attached: [Screenshots of the 167-hour limit and the conflicting quota bar]
The only option I am now using is to keep working till it gets to 20%, then use the Flash model. For small works, I use the Flash model, and for complex works, I use the High Thinking model. Now I am not using any Opus or other models. I don’t know when this nightmare is going to end.
This has been a persistent issue for me, as well. I had read that if you do not completely deplete the available rate limit during the 5 hour window, that it will properly reset. At first, I thought this suggestion was effective as I utilized about 40% of the 5-hour allotment, and it appeared to properly reset to 100% at the 5-hour mark. However, during the next window I utilized 80% of the allotment, and at the 5-hour reset point the tracker (Antigravity Cockpitv2.1.20 ) updated to reflect that I had 20% remaining that would not reset for 5 days 20 hours and 43 minutes. So… I don’t know. If Google had a better coding model, this wouldn’t be as big of a deal… but they don’t, so it is.
The Situation: I am a subscriber to the Google AI Pro (2 TB) plan. The explicitly stated policy for model usage (specifically for Claude Sonnet 4.5 / Opus 4.6 on Antigravity) is a reset cycle of 5 HOURS.
Evidence: (Please see the attached screenshot showing the 167h 52m reset timer on a Pro account)