Google Modifies “Thinking” Mode Access in Gemini 3 Pro, Cites High User Demand

Google has reportedly adjusted the free access limits for its advanced Gemini 3 Pro AI model. The changes, which affect the “Thinking” mode, have been introduced due to increasing demand for the AI platform. While precise limits have not been disclosed, the updated support page notes that daily access for free users may change frequently. The adjustments appear to be applied mainly in the United States, and similar updates for users in India have not yet been confirmed.

Before the launch of Gemini 3, free users were allowed up to five prompts per day, consistent with the previous Gemini 2.5 Pro limits. The new system is based on a token-based model, in which the complexity of queries determines daily usage. Simple prompts may consume fewer tokens, allowing multiple queries, while more complex requests could quickly reach daily limits. This dynamic system ensures that AI usage is balanced according to the computing resources required.

In a related update, Google has temporarily rolled back certain features of NotebookLM, its AI-powered research platform. The Infographics and Slide Decks features have been made unavailable to free users, while additional limits have been applied to subscribers of Google AI Pro plans. Google cited “overwhelming demand” and “capacity constraints” as reasons for the rollback. The company has indicated that these features will be restored in the future.

The changes are expected to encourage more efficient use of AI resources and maintain service quality amid growing adoption. Free users continue to benefit from access to Gemini 3 Pro AI for tasks such as research, content creation, and learning assistance. Students, in particular, can leverage the AI for project work, homework support, and study assistance. When NotebookLM features are restored, users are expected to gain enhanced tools for presentations, infographics, and organized research.

These updates demonstrate Google’s effort to balance the increasing global demand for AI services while ensuring that both free and paid users receive optimal functionality. The AI’s adaptive token-based system allows for flexible use based on query complexity, supporting both casual and advanced applications.