New Delhi: Google has recently revised how usage limits work for its Gemini 3 AI models, removing the shared quota between the Thinking and Pro models and increasing daily prompt allowances for paid users. According to some reports, each reasoning-focused model will now have its own independent usage limit, making it easier for users to choose the right model without worrying about exhausting a shared pool. The change applies across the subscription tiers and is aimed at improving transparency and control when switching between Gemini models for different tasks, such as problem-solving, coding, or advanced reasoning.
Google currently offers three options in the model picker for the Gemini 3. The Fast model is especially designed for quick responses, while Thinking focuses on solving complex problems efficiently. The Pro model is meant for tasks that require deeper reasoning, such as advanced mathematics and coding. Thinking and Pro models shared the same daily prompt quota. This meant that the heavy use of one model would reduce the available prompts for the other.
This setup may have caused confusion, especially for the users who wanted clear boundaries when selecting models for specific tasks. The company has now unveiled separate limits for each model. For AI Pro subscribers, the daily allowance has been set at 300 prompts for the Thinking model, while the Pro model is set at 100 prompts per day. For AI Ultra subscribers, the Thinking model limit has been raised to 1,500 prompts per day, with the Pro model allowing 500 prompts per day.
Google has also confirmed that these independent limits also apply to free users, although the specifics have not yet been disclosed. In addition, access to the Thinking and Pro models remains more restricted compared to paid plans. The Fast model continues to offer broader general access. The Google AI Pro subscription starts in India at Rs 1,950 per month, and the AI Ultra subscription starts at Rs 24,500 per month.
