AI Pro - Antigravity Model Quotas changed spontaneously?

I recently upgraded to Google AI Pro this past week, and the core reason was the improved quotas refresh timers within Antigravity, going from 5-7 days to 5 hours.

I was working on some projects late last night with no quota issues, but after wanting to get back into them today, the limits for Gemini 3.1 Pro have increased from 5 hours to 5 days… I’ve tried signing out and back in, disconnecting third-party connections and reconnecting, age verification (as this was an area another user recommended trying to fix a potential quota loop), but nothing has worked.

Anyone else experiencing this? Pretty disappointing and I’m honestly looking to spend my money elsewhere following this. What was the point in upgrading if things are adjusted without communication or notice?

2 Likes

Same for me. I’m on the Pro plan, and today, after the Gemini 3.1 model spent two hours trying to fix ten lines of code without success (I was trying to make it fix the bug out of curiosity—just exploring the model’s capabilities), I decided to check my quota and found that, although my 3.1 limits are not yet exhausted, they will be refreshed in 1.5 days. Yesterday, quota refreshes were supposed to occur every 5 hours

same problem here my quota will reset in 6day.

This is a real deal breaker… Quotes suddenly exhaust so fast and refresh in 4 days and I´m also on the Pro-plan. And the update to use 3.1 broke my AI chat history…

1 Like

Pro tip: read the forum before posting. Had you done so, you’d have the answer.

1 Like

I see you’re going around posting this useless comment in every thread. Let me guess - do you work for some Google competitor and your goal is to drive customers away from here?

Thanks Steve, much appreciated bud.

Nope. I’m an AG user who’s annoyed by all the whining about the quotas, and the repetitive questions that are clearly answered.

1 Like

Now I understand who AI is going to replace - people whose IQ isn’t high enough to think of adding a link to a post with an explanation.

Subject: Recurring Quota Regression: 3-Day Reset vs. Advertised 5-Hour Cycle (Second Occurrence)

I am reporting a recurring anomaly regarding Gemini 3.1 Pro quota management. This is the second time my account has spontaneously shifted from a 5-hour rolling reset to a staggering 3-day lockout.

Observation:

  • Gemini 3 Flash: Operates normally on a 5-hour cycle.

  • Gemini 3.1 Pro: Suddenly displays a “3d 8h 59m” reset window.

As a Pro subscriber, a 15x increase in reset latency effectively renders the model unusable for active development. This inconsistency between models suggests a backend configuration error rather than a deliberate policy. Please investigate why this regression persists for paid tiers.

Yes, I’m on pro also, and from 4 hours, gradually claude move to several days, and now all models align with claude, from 4 hours to over 130 hours. It makes pro tier not workable for me.

I’m on the Pro plan and the same thing happened to me – my Gemini 3.1 Pro quota suddenly switched from the advertised 5‑hour refresh to a 3‑day reset window without any warning.

One thing I’ve noticed: the exhausted quota / long reset window is only affecting Claude 3.5 Opus for me – Gemini Flash and Gemini Pro are still working normally with their usual limits.

That makes this even more confusing, because it feels like a Claude‑specific quota change rather than a global Pro account limit.