Google boosts Gemini coding quotas ninefold after backlash

Google has significantly increased usage limits for its Gemini AI models within its Antigravity coding environment after a swift and widespread backlash from paying users. The company implemented the changes over several days in May 2026, first tripling and then tripling again the weekly quotas for all paid Antigravity plans, resulting in a net ninefold increase from the levels that had sparked complaints.

The controversy began when Google introduced stricter, compute-based usage limits for Gemini models following its annual developer conference. Users of the paid AI Pro plan, particularly those relying on Antigravity for complex coding tasks, reported hitting their new weekly allowances within hours or after just a couple of work sessions. Developers expressed frustration on forums, describing the sudden restrictions as a barrier to productivity for a tool they depended on for lengthy workflows.

In response, Varun Mohan, a director at Google DeepMind working on Antigravity, announced the initial adjustment, tripling rate limits and resetting weekly quotas. Days later, he announced a second, identical adjustment. Mohan acknowledged that the company saw a huge jump in users...