Google has enhanced NotebookLM by introducing a new reasoning engine, additional file output choices, and a more adaptable research process. This gives the AI notebook a wider range of features for managing intricate projects.

Enhanced Reasoning with Gemini 3.5

The update transitions NotebookLM to Gemini 3.5 and Antigravity, Google's coding model. This results in more precise answers and increased insight into the tool's decision-making process.

Google indicates that every notebook operates on a dedicated cloud computer capable of writing and running code, supported by more than 100 specialized software skills. The company's benchmarks show the updated system achieving a 65 percent-plus win rate compared to the prior version in five main evaluation areas, with notable improvements in analyzing large documents at 69.9 percent and web research at 78.2 percent.

Expanded Output Options and Easier Starting Point

NotebookLM now produces a broader selection of downloadable files from the studio panel. These include PDFs, Word documents, Excel spreadsheets, PowerPoint presentations, CSVs, data visualizations, and images. Users can specify detailed formatting guidelines before creation and ask for modifications later.

For research, NotebookLM does not require a complete source collection to begin. Users can start a notebook with a basic concept or query, and the tool will leverage Google Search to identify pertinent sources and assist in developing the repository. Users maintain complete oversight on which sources are included, and all sources stay properly credited within the notebook.

The updates are available immediately for Google AI Ultra subscribers and Workspace business customers on AI Ultra Access and AI Expanded Access plans, with wider rollout expected soon.