Google has adjusted how usage limits work for its Gemini models, a change that quietly alters the experience for people who rely on the system for extended or technical sessions. The update separates daily usage caps for the Thinking and Pro models, ending the shared quota that previously applied across both. For users who frequently switch between complex reasoning tasks and more traditional coding or math prompts, the revision removes a long-standing constraint.
Until now, both Gemini models drew from the same daily allowance. A prompt used to reason through a multi-step problem would reduce the same pool that covered code generation or advanced calculations. That structure often forced users to ration prompts or avoid certain tasks late in the day. According to reporting by 9to5Google, Google has now split those limits and raised them independently for paid subscribers.
On the Google AI Pro plan, the Thinking model is now capped at 300 prompts per day, while the Pro model remains limited to 100. The higher-tier Google AI Ultra plan increases those numbers significantly, offering 1,500 daily prompts for Thinking and 500 for Pro. Users on the free tier still have access to both models, though Google continues to describe that access in general terms, noting that daily limits may change without notice.
The practical impact is straightforward. The Thinking model, designed for longer chains of reasoning and more complex problem-solving, no longer competes directly with the Pro model’s allowance. Someone working through a detailed analytical task can do so without worrying that it will reduce their capacity for code generation or math-heavy prompts later in the session. While the Pro model remains the primary option for structured outputs like programming tasks, each now operates within its own defined boundary.
This adjustment also reflects the realities of managing demand. Gemini 3 has seen steady uptake since launch, and Google has periodically tightened or relaxed limits on the free tier, particularly around image generation, when usage spikes. By increasing and separating caps for paid plans, the company appears to be aiming for more predictable access rather than simply expanding overall capacity.
The change does not introduce new features, nor does it alter how the models behave. Instead, it addresses a friction point that emerged as users pushed Gemini into longer, more intensive workflows. For subscribers, the result is fewer trade-offs during daily use and a clearer sense of how much work can be done before hitting a ceiling. For Google, it is another incremental step in balancing infrastructure limits with growing expectations around always-available AI tools.
