Google updates NotebookLM with five features for research, presentations and creative work

Google has significantly expanded NotebookLM: the service has moved beyond its role as a "smart notebook" and become a multimodal environment for researching, structuring and packaging ideas. For product teams, researchers, designers and writers, five features matter most right now, because together they cover almost the entire path from raw material to a finished presentation.

Research without the grunt work

The main change is Deep Research mode. NotebookLM used to answer mostly from files the user uploaded. Now it can search the internet for new sources on its own, cross-reference them and build reports backed by citations. This changes the early phase of any project: instead of collecting links by hand, you can build a working corpus of materials inside a single notebook right away. In effect, the tool starts to act as a research assistant, not just an interface to documents you have already gathered.

The practical upshot is that research stops being a separate stage. Web sources it finds immediately become part of the shared context that the chat, mind map and content generation then draw on. If you weed out weak material and adjust the search focus, NotebookLM gradually builds a high-quality knowledge base for a specific task, whether that is a product strategy, a creative concept or a client brief. This matters most where data changes quickly and a team needs a current picture without dozens of tabs and manual copy-pasting of notes.

Mind maps and presentation

The second strong block is visualizing structure and turning it into a clear narrative. The Mind Map feature automatically shows topics and the links between them, grouping fragments from sources into navigable nodes. That helps where linear text makes it hard to see the whole system: you can quickly spot blind spots, overlapping constraints, or topics that have so far been explored only superficially. From the map, you can jump to the branch you need and expand it into a separate working scenario.

The same knowledge base then becomes material for external communication. In Studio, NotebookLM generates infographics and slides, and recent updates added prompt-based editing of presentations and PPTX export. Audio Overview and Cinematic Video Overviews work separately: the service produces podcast-style audio overviews and videos with a set tone and pace, so you can check the logic of a story faster before a meeting, workshop or internal demo.

• Deep Research for finding and vetting new sources
• Mind Map for navigating topics and connections
• Studio for quickly assembling slides and infographics
• Audio and Video Overviews for testing delivery and rhythm
• PPTX export for handing a draft off to PowerPoint

One notebook for everything

Another important update is the scale of the workspace itself. NotebookLM, now built on Gemini 3, gets a context window of 1 million tokens and support for different input types, including Word documents, spreadsheets and OCR scans. That means you can keep almost the entire history of a project in one place: research, timelines, annotated diagrams, interview notes and long transcripts, without trimming material to an artificial minimum. For product and architecture work, this removes one of the main problems of LLM tools: you no longer have to keep deciding which files to leave out of the context.

For complex decisions, Data Tables are especially useful. The service can turn qualitative descriptions and scattered observations into structured comparison matrices, which can then be exported to Google Sheets. The format is handy when you need to lay out options against criteria, compare competing concepts, or show stakeholders why one hypothesis looks stronger than another. It fits particularly well with choosing product directions, partners, rollout scenarios and the requirements of different teams.

"This is no longer just an app for synthesizing data, but a hub for

designing complex creative systems."

What this means

NotebookLM looks less and less like a supporting AI notebook and more and more like an end-to-end workspace for knowledge and storytelling. If Google keeps up its pace of updates, the service could become a standard tool for teams that need not only to read and summarize materials, but also to quickly turn them into a decision map, a presentation and a clear story for the business. For anyone working with a complex body of knowledge, that saves hours on every cycle.