Gemini Notebook, previously called NotebookLM, is Google’s AI-powered research and synthesis tool. You upload documents and it becomes an expert on those documents specifically, answering questions, drawing connections, and generating summaries grounded in your selected sources. Google renamed NotebookLM to Gemini Notebook in July 2026. It remains available as a standalone product while also offering higher limits through Google AI and qualifying Workspace plans.
Grounding is the defining characteristic. Unlike a general-purpose chatbot answering from broad training data, Gemini Notebook bases its responses on the sources selected for the notebook and cites the relevant passages. If the answer is not supported by those materials, it should say so. This makes the tool especially useful for research requiring accuracy and traceability, although its AI-generated responses still need verification.
Supported source types are impressively broad: PDFs, Google Docs, Google Slides, text files, Markdown, web URLs, YouTube links, audio files, and EPUB files added in March 2026. Upload a podcast episode, a set of research papers, a competitor’s website, and your own strategy document simultaneously, and Gemini Notebook can synthesize across them in response to a single question.
Audio Overviews are the breakout feature that drove mass adoption. Gemini Notebook generates a two-host AI podcast discussing the content of your uploaded sources, with natural conversational flow, interruptions, and illustrative examples. On the free plan, you get three per day. It is genuinely useful as a learning tool: upload a textbook chapter, generate a podcast, and listen during a commute.
The March 2026 update added Cinematic Video Overviews, slide deck generation with prompt-based per-slide editing, PPTX export, ten new infographic styles, improved flashcards with progress saving, and saved conversation history. These are not surface features. They turn Gemini Notebook from a Q&A tool into a content production system built on your own documents.
Collaboration is more capable than it first appears. Users can share notebooks, and editors can modify shared notes with real-time synchronization. However, the experience remains less complete than Google Docs for team research because it lacks the same depth of commenting, simultaneous document editing, and collaborative review controls.
Export and integration options still have limits. Notes can be exported directly to Google Docs or Google Sheets, but there is no public consumer API or broad webhook and integration layer for automating downstream workflows.
The free plan is sufficient for many individual users: 100 notebooks, 50 sources per notebook, 50 chats per day, and three Audio and three Video Overviews daily. Google AI Plus raises those limits to 200 notebooks, 100 sources, 200 chats, and six Audio and six Video Overviews per day. Google AI Pro adds 500 notebooks, 300 sources, 500 chats, 20 Audio and 20 Video Overviews, and two Cinematic Video Overviews per day. The two Ultra tiers increase the limits further, including ten or 20 Cinematic Video Overviews per day.
