AI Model Quota Limits Have No use from Now It getting worst day by day

As a Pro plan user, I wanted to share some feedback regarding the recent Antigravity 2.0 update and the current quota system.

When I first started using Antigravity, the experience was genuinely great. The model limits felt reasonable, and the Flash model provided a practical balance between performance and quota usage. It allowed longer and more consistent workflows without constantly worrying about hitting limits.

However, after the recent updates, quota consumption has increased significantly. Even when using the Flash model, the quota now drains extremely quickly. With the Antigravity 2.0 changes, both Flash and Pro models now appear to share the same quota pool, which removes the main advantage Flash previously had for lightweight or iterative tasks.

The refresh system has also become difficult to work with. Earlier, limits would refresh in around 5 hours with partial recovery, which was still manageable. Now, after that partial refresh, users are pushed into a 7-day waiting window similar to Pro and Claude limits. At that point, it becomes very hard to rely on the platform for consistent development work.

I want to clarify that I do not think Antigravity itself is bad. In fact, the product and model quality are still very good. The issue is primarily with the usage limits and credit system, which now feel far too restrictive for practical daily use.

For example, I tested a very small single-change request using the Flash model with credits enabled, and it consumed around 50 credits. If that scale of usage is accurate, then the monthly allocation of 1000 credits does not seem sufficient for serious or continuous workflows.

Because of this, I am considering moving to alternatives like Codex or Cursor, mainly because they currently offer limits that are more practical for long-term usage.

I hope the team reconsiders the current quota structure, especially for Flash usage, because the platform previously had a much better balance between usability and limits before the latest update.

3 Likes