Google has announced a new update for NotebookLM, introducing major performance and personalization improvements. The update enhances context understanding, conversation memory, and user control, powered by the latest Gemini models.
Upgraded chat and deeper insights
Google has improved NotebookLM’s chat system with new back-end upgrades that enhance response quality and contextual accuracy. These Gemini-driven changes deliver more reliable results, especially when referencing multiple sources. In internal testing, the company observed a 50% increase in user satisfaction for responses that used large data sets.
Expanded context and multiturn conversation capacity
The update activates Gemini’s full 1 million-token context window across all NotebookLM plans, allowing it to handle larger document collections. Conversation capacity has been expanded over sixfold, enabling NotebookLM to maintain coherent and relevant exchanges across extended sessions.
Improved information synthesis
NotebookLM now automatically explores source material from multiple angles to generate more comprehensive and well-grounded answers. This capability helps users discover new connections between documents and ensures results remain contextually accurate, particularly in large or complex notebooks.
Saved and secure conversation history
All conversations in NotebookLM will now be automatically saved, supporting users who work on long-term research or writing projects. Chat sessions can be resumed later without losing context, and users can delete conversation history at any time. In shared notebooks, chat logs remain private to individual users.
Goal-based chat configuration
Google has introduced goal-based customization for all NotebookLM users, allowing them to define how the chat behaves. This includes setting specific roles, tones, or objectives suited to different workflows.
Examples include:
- Acting as a research advisor that questions assumptions and reviews arguments.
- Serving as a marketing strategist focused on clear, actionable steps.
- Analyzing material from academic, creative, and critical perspectives.
- Functioning as a Game Master for text-based simulations with user-directed choices.
Anuja Agrawal, Senior Staff Lead for NotebookLM, said the integration of the latest model improvements enables more precise and context-aware responses. Agrawal noted that customizable goals help NotebookLM adapt better to varied user needs and project types.
Availability
The saved and secure conversation history feature will begin rolling out to users over the next week. Other performance and personalization upgrades are now available globally across all NotebookLM plans.