Antigravity quota for google ai pro user

Hello Antigravity Support,

I’m writing to express my frustration regarding the usage limits on my paid Google AI Pro subscription.

I purchased Google AI Pro expecting reasonable access to the premium AI models. However, my experience has been extremely disappointing.
The Claude and GPT models became unavailable after only about 20 minutes of use. I have been locked out of the Gemini models for 3 consecutive days, continuously waiting for the limits to reset. Today, it finally shows they will be available again in about 2 hours, meaning I’ve effectively been unable to use one of the main features I’m paying for. Even worse, after only around 20 minutes of using the Claude/GPT models, I’m now seeing the message: Your weekly limit will fully refresh in 5 days, 21 hours.

As a paid Google AI Pro user, this is unacceptable. These limits are so restrictive that the subscription provides very little practical value. I also completed the required age verification, so there should not be any additional restrictions affecting my account.

I would appreciate an explanation for:

  1. Why the weekly limits are being exhausted after such minimal usage.
  2. Why the Gemini models have remained inaccessible for three straight days.
  3. Whether there is an issue with my account or if these limits are being applied incorrectly.
  4. What can be done to restore the level of access that a paid subscriber should reasonably expect.

I subscribed in good faith, expecting reliable access to the AI models included in Google AI Pro. Instead, I have spent more time waiting for limits to reset than actually using the service.

Please investigate this issue as soon as possible and provide a solution.

Thank you.

The crux of the problem is still what you want the AI to do. You just have to read the forum posts to see that many people complain about the limits, but hardly anyone talks about what they actually did. And the fact is that a limit worth $20 or $25 will pay out exactly that amount within a given time window. Google itself makes it difficult to track usage by stating that a given plan offers X times more than the previous one, but we don’t know exactly what the limit was for that previous plan. So, if a given plan cost a certain amount, we get exactly what we paid for, but we can use that limit several times over the course of a month—which is different from using the free version. Because, as we know, you can use AI without paying.

Jusst wait and wait then "working… " and bomb ~!~~ out off token ???

what i can tell from my experience using the pro plan for a few months now is:

  1. they keep changing “how many quota” the user have without even telling us how many tokens we have used and what’s the actual token limit, it’s just an arbitrary concept of “you have used this much”
  2. previously we signed for an AI tool that have individual pool between family members, then they changed it when antigravity 2.0 was released, and the quota… i can tell that it is even lower than the previous one as i keep getting weekly limit
  3. what’s worse is gemini 3.1 pro (high) is not as smart as claude opus or sonnet (at least from my work experience), and with more limit to the gemini quota while they keep making mistake (a lot) even after i laid a pretty solid plan with opus 4.8 (high), i 100% wont extend my google AI subscription, as my previous consideration was i plan with opus, and implement with gemini.

and this is just after a short session only today, i already used up 9% of my weekly quota. a very short session as today is sunday and i sparingly sit to work.

I wonder why google takes us for fools and shadow decreases quotas without notifying users, like it or not is it Noticeable

That’s all true, and it should also be added that models with cheaper inputs and outputs are being removed from general access, forcing people to use expensive models for simple tasks. For example, Gemini 3 Flash from Antigravity, where this model was unlimited in the Pro plan.

exactly.

I don’t know what they are thinking, the only advantage Google has over Anthropic is their massive context size and seemingly unlimited quota for small task or plan implementation.

now that they’ve taken that over, I’d rather subs 2 Claude pro than 1 Claude and 1 Gemini for my personal project’s AI.
(since I got Claude max from my company for work related stuff, not personal)