The Latest Update to NotebookLM Makes It an Even Better Research Tool.
NotebookLM is a lesser-known, research-focused tool from Google that complements Gemini, and it’s received a significant update that makes it even more useful. If you need to delve deeper into research, exploration, reporting, or analysis, it’s one of the best tools available .
Here’s what’s new in the latest update: NotebookLM can now write code (via Google’s Antigravity development platform) and output results in a variety of file formats, including PDF, PNG/SVG diagrams, Excel spreadsheets, and PowerPoint presentations. You can also edit these files after they’ve been created by following the prompts.
If you don’t have any sources to begin with, you can discuss it with NotebookLM, and the program will help you find suitable sources online—right during the conversation. This is a more natural approach to the topic if you’re starting from scratch.
The AI will show you its reasoning as it unfolds, so you’ll receive regular updates on what the AI is doing or thinking, rather than just icons with spinning dots. In fact, it’s much more like the standard Gemini app, while maintaining the focus on exploration and learning.
These updates are already available to users of the Google AI Ultra plan (costing $100 or $200 per month), and Google says it plans to roll them out to “other users over time,” without providing details. I was able to test the new features using the Google AI Ultra plan.
Getting Started with NotebookLM
The traditional way to get started with NotebookLM is to upload sources (PDFs, web links, YouTube videos, and so on) and then start asking questions about the collected information. Now, you can immediately engage in a conversation with the AI-powered bot and obtain additional sources along the way.
To test this in practice, I decided to see what NotebookLM could find online about Christopher Nolan films. The AI tool not only found a wealth of relevant information but also synthesized it into a useful overview. An “Import” button appeared at the end of the overview, allowing sources cited by NotebookLM to be added immediately. Furthermore, the original answer is also added as a source.
Previously, NotebookLM could search for sources online independently, but this process was split into several separate interface panels and required several clicks to import all the data. Now, it’s much more conveniently integrated into the main chat window, allowing you to begin research and analysis more quickly.
You can still add sources manually via the left panel, and as always, NotebookLM will cite these sources in its responses. If you’re ever curious about the origin of an idea or fact, you can click the “Cite” button to find out. In this case, Wikipedia did most of the work.
While I haven’t noticed any hallucinations here, a strange situation has arisen: a search for a list of “key” actors from Nolan films yielded numerous men, but only two women from his 12 films: Marion Cotillard and Anne Hathaway. Apologies to Carrie-Anne Moss, Hilary Swank, Katie Holmes, Scarlett Johansson, Rebecca Hall, Elizabeth Debicki, and (Oscar-nominated) Emily Blunt.
I think this is because NotebookLM is fixated on Nolan’s Wikipedia list of frequent collaborators, but that doesn’t fully explain the situation. It’s worth keeping this in mind if you rely on AI in your research: even if actual errors are rare, the process of generalization and comparison these tools perform still needs to be verified.
NotebookLM does a great job of creating files.
Following several additional requests to increase the representation of women in this Nolan film, I decided to test another feature: the ability to create editable files. You can request Microsoft Office files, PDFs, and PNG or SVG diagrams and receive a downloadable file that you can open.
I asked for a PowerPoint slideshow covering all of Nolan’s films, and NotebookLM delivered: I received a well-formatted, typo-free introduction to each film, including the plot and themes of each, as well as the key actors. NotebookLM still can’t generate images or retrieve them from the internet, so you’ll have to create that yourself.
NotebookLM now has another feature: requesting edits to existing files, similar to Nano Banana . For example, for a PowerPoint presentation, I requested changes to the fonts and background color, and NotebookLM followed my instructions precisely, creating a shareable slideshow in just 10-15 minutes, whereas otherwise it would have taken me at least a couple of hours.
Next, I tried creating a PDF guide to Nolan’s films, and while the design certainly isn’t award-winning, it turned out quite well-structured, clean, and consistent. The information was accurately sourced from Wikipedia and the British Film Institute , and this again really saves time—it seems the days of copying, pasting, and reformatting answers from NotebookLM are a thing of the past for many tasks.
NotebookLM has a well-implemented feature for switching between landscape and portrait orientations for PDF files, and while making minor edits—such as moving a line or adding a new text box—is more difficult without a visual interface, you can always make these changes manually if necessary. Or, if you’re editing an Office file and want to fully utilize artificial intelligence, Copilot is happy to assist you .
I didn’t try Antigravity’s new programming features here, as that’s a completely different topic and I have little experience with it. As for the other new features, NotebookLM is clearly better and more useful than ever, even if the AI’s age-old warning about double-checking your answers still applies.