NotebookLM Agentic AI Update Is HUGE! Agentic Coder Now?

Key Takeaways

Looks like Notebook LM just got an insane amount of new features as it's evolving from a simple chat with your document tool into an agentic research assistant. Now, it can even act like a lightweight coder. Notebook LM can reason across resources, handle multi-step research tasks, generate structured outputs, analyze data, create charts, and build full research artifacts. The biggest upgrade is the agentic research feature. Instead of only answering questions from files you upload, it can now help you discover sources, reason through topics, and it can work through complex research problems. Now, right now this is a feature that is rolling out with Google AI Ultra subscribers, but availability will be coming for all paid plans soon. First up is the chat experience which has been completely upgraded. Notebook LM is now powered by Gemini 3.5 Flash and Antigravity, which means you get more thoughtful AI experiences with better visibility into the how the model is actually thinking throughout the whole process. But, the craziest part is that each notebook now comes with a secure cloud computer. This includes over 100 curated software skills. So, instead of summarizing PDFs or answering questions from your sources, Notebook LM can handle deeper research or complex analysis and multi-step workflows directly inside your notebook. This is basically Google turning Notebook LM into a full agentic research workspace. You upload your sources, ask it to investigate something, and now it can reason tools and cloud environments that actually work through the task in a much more advanced way. If you want the best AI tools, workflows, and drops before everyone else, join my free newsletter with the link in the description below, which is completely free. And this is a pretty big deal because Notebook LM is slowly turning into more than just a place where you dump PDFs and ask questions. With the code execution, web search, source directory, and a secure cloud computer built into each notebook, it's going to start feeling more like a mini research agent that can actually do work for you. So, instead of just asking summarize the document, you can ask it to compare multiple sources, find gaps in research, analyze data, generate charts, organize findings, and help you understand what actually matters. And the fact that Google is baking this directly into Notebook LM is definitely insane. Now, in terms of the performance boost by having the Gemini 3.5 and anti-gravity power all of the notebooks, they have actually showed a couple of early benchmark numbers, and this is where this is a system that beats the previous system over 65% of the time across the top five evaluation areas: accuracy and quality, multilingual support, large document analysis, document creation, and advanced research. And you can see a substantial gain in comparison to the previous system with the newer one. For large document analysis, it had a 69.9% win rate. And for advanced web search research and source directory, it reached a 78.2% win rate against the old baseline. So, this is not just a small quality of life update. Google is clearly trying to turn Notebook LM into a real research agent that is powerful. Next is probably one of the best features I have seen, which is where Notebook LM can turn your sources into actual downloadable outputs. So, instead of just chatting with your documents and getting the text answers, you can now ask Notebook LM to assemble the content from all of your sources and then turn it into proper files, which is great cuz this includes data visualization and charts being outputted into PDFs, doc files, you have markdown documents, Excel sheets, PowerPoints, CSVs, JSON files, and even images using Nano Banana. And this is actually a massive upgrade because it turns Notebook LM from a research assistant into something that can help you create a fully final deliverable with these different sorts of file outputs. You can give it business documents and ask for a PowerPoint. You could give it raw information and ask for charts, spreadsheets, or clean structured data. So, the workflow is no longer just asking questions and copying answers. This new agentic update makes it more intelligent and gets you better outputs. And like I had stated at the start, Notebook LM is turning more into an agentic researcher companion. This basically means you're no longer having to start with a perfectly organized notebook full of sources. You can literally start with loose ideas, random questions, or rough research directions. Notebook LM is going to be intelligent enough with this agentic loop to help you guide through all of the building blocks for your entire notebook. The craziest part is is that it can even find and suggest relevant sources from the web. It can then add them to your notebook with your permission. So, this changes the workflow completely cuz instead of just doing all the source hunting, selecting links, uploading files, and then asking questions, and after that, Notebook LM can help you build a research base from scratch. You give it an idea now and it helps you figure out what sources you need. It can discover these sources and then helps you analyze everything inside one workspace. An important part here is that Google is making the research process way more intimidating. Before Notebook LM was mostly useful when you already had your PDFs, links, notes, and sources ready to go. But now, it can actually help you shape the research direction from the beginning. So, if you're trying to understand a topic from different perspectives, finding primary sources in another language, or discovering related work from an author you just found, this is an update where Notebook LM can help surface those sources for you And and even though it uses Google search to find high-quality web sources, you can still stay in control. This is where it won't just randomly dump sources into your notebook. It's going to ask you for your permission and then it's going to be attributed into your notebook after you have basically specified that you want to have this source added to your notebook. For researchers, this means you can bring messy data from different sources and have Notebook LM clean it up for you. You can research extra content, run analysis, and then turn everything into charts and even a final report that you can even download. When it comes to technical teams, this is where you could have it useful for complex documentation, where it can now be transformed into something way easier to understand and this is where coding kind of comes in, where you can research upon different sorts of newer documentation for different languages or different packages and then have it fully condensed into a markdown report, which you can download and then export it and then send it over to your coding agent. Another way you could use this is instead of just being a program manager, where you manually read through all of the dense specs, you can essentially have it so that Notebook LM can turn in that all of that into simplified guides, a road map and even a slide deck. For small businesses, you can use it as a way for you to upload campaign data, sales numbers and even ad spend and then you can essentially have Notebook LM calculate what actually worked and whether the campaign was worth scaling. So overall, it's not just about adding more formats or smarter chats, it's about getting the best usage out of Notebook LM by having it become more authentic. Another cool feature that they're adding is source attribution for artifacts. So now when Notebook LM generates a report, chart, slide deck or any output, you can see exactly what prompts and sources were used to create it. That's huge for trust because you're not just getting a random AI-generated file. You can actually trace where the information came from and see what you can actually do with the sources. If you like this video and would love to support the channel, you can consider donating to my channel through the Super Thanks option below. Or you can consider joining our private Discord where you can access multiple subscriptions to different AI tools for free on a monthly basis, plus daily AI news and exclusive content, plus a lot more. Also, this is more speculation, but based off of what I have seen from internal testers, as we all know, Notebook LMs' video capabilities right now are not the best. It's still pretty good in most cases, but soon they're going to be adding the new Omni video generation model from Gemini into Notebook LM included in your plans. And that way you're going to be able to get better video generations based off the notebook that you have. So, think about it. You can create some great infographic videos or even some beautiful animations based off your notebooks. But overall, this was a huge update by Google, and they have essentially made Notebook LM far more authentic and more capable. I'll leave all the links that I use in today's video in the description below. Let me know what you guys think. Make sure you go ahead and take a look at the second channel, join the newsletter, join the Discord, follow me on Twitter, and lastly, make sure you guys subscribe, turn on the notification bell, like this video, and please take a look at our previous videos so that you can stay up-to-date with the latest AI news. But with that thought, guys, have an amazing day. Spread positivity, and I'll see you guys fairly shortly. Peace out, fellas.