Gemini Notebook Learns to Build Interactive Websites with Flashcards from Your Notes
Google has taught Gemini Notebook (formerly NotebookLM) to turn notes and uploaded sources into an interactive website with flashcards and diagrams. A ZDNET reviewer asked it to explain cellular respiration stage by stage and within a few minutes received a tabbed site whose answers draw on her sources. The feature is already available in Notebook, with the mobile version coming later. For studying and internal knowledge bases, it works as a cheat sheet that requires no code.
AI-processed from ZDNet AI; edited by Hamidun News
Anyone who has studied for an exam or tried to make sense of someone else's documentation knows the pain: there is plenty of material, it lies scattered around, and turning it into a clear structure falls to you by hand. Google wants a neural network to take on that work. In Gemini Notebook, formerly called NotebookLM, a feature has appeared that turns notes and other sources into an interactive page, or essentially a small website.
What Exactly Has Appeared
Gemini Notebook is built around materials you upload yourself: textbooks, files, study notes. It has three tabs: Sources, Chat, and Studio. The new capability lives in Studio, in the list of what you can create from your sources. The result looks like a set of themed pages with self-check flashcards, dashboards, and colored panels. The idea is that complex information can be explored piece by piece rather than read straight through.
Google announced the update on Tuesday. The feature is currently available to Notebook users, and it will reach the mobile app later. According to ZDNET, the company has not announced a timeline.
How It Worked in Practice
ZDNET reviewer Maria Diaz tested the feature on a study task. She had no study materials of her own, so she uploaded several online sources to the notebook. She then gave the prompt: build a site that visually explains cellular respiration, walks through each stage with interactive diagrams, and includes flashcards for self-testing.
Gemini first cited the sources in the notebook, then spent several minutes assembling the page. After that it reported that the site was ready, briefly summarized what it had found in the sources relevant to the request, and pointed her to where the result was. According to the author, the outcome was complete and tidy: the information was laid out across tabs, and the site was easy to navigate. She liked the flashcards best: hovering over them shows which ones you have already gone through.
Sourcing matters here. The site is built from what sits in the notebook, not from the model's general knowledge. For study tasks, this lowers the risk that an invented fact ends up in the cheat sheet. But it does not guarantee full accuracy: testing one request about school biology does not prove accuracy on complex or contradictory material.
Who Will Benefit and What the Caveats Are
The most obvious scenario is studying. A student uploads lectures and a textbook, gets an interactive study guide, and shares the link with classmates. The second scenario is closer to work teams: from a set of internal documents, you can quickly assemble an understandable introductory primer for newcomers. The third is product and research notes, which are easier to show colleagues as a navigable structure than as a long block of text.
There are also limitations worth keeping in mind. First, these are the impressions of a single reviewer, someone who, by her own account, knows little about programming. For her, the main advantage was that she did not have to write any code. For developers, such a site is unlikely to replace proper front-end work: ZDNET reports no ability to finely adjust the design and no option to export the source code. Second, the material contains no data on limits, pricing, or how the feature handles large volumes of sources. Those details will need to be checked independently.
The practical takeaway is simple. If you already keep your materials in NotebookLM, that is, in Gemini Notebook, open Studio and try turning one of your notebooks into a site. Check how closely the answers match your sources, and only then send the link to other people. For quick cheat sheets and training materials, the tool looks useful; for serious publication, the output is better verified by hand.
- Nov 21, 2026Нейросеть MIRA рисует Rocket League в 20 кадрах в секунду, но пинг доходит до 460 мс
- Nov 21, 2026Google Opens Pre-Orders for Gemini Smart Glasses from $395, Shipping December 10
- Nov 21, 2026Google to split Gemini chat history from model training, so you can opt out without losing your history
- Nov 21, 2026Gemini Notebook Can Now Build Interactive Websites from Your PDFs and Notes
Key Questions Answered
What Exactly Has Appeared?
Gemini Notebook is built around materials you upload yourself: textbooks, files, study notes. It has three tabs: Sources, Chat, and Studio. The new capability lives in Studio, in the list of what you can create from your sources. The result looks like a set of themed pages with self-check flashcards, dashboards, and colored panels. The idea is that complex information can be explored piece by piece rather than read straight through. Google announced the update on Tuesday.
How It Worked in Practice?
ZDNET reviewer Maria Diaz tested the feature on a study task. She had no study materials of her own, so she uploaded several online sources to the notebook. She then gave the prompt: build a site that visually explains cellular respiration, walks through each stage with interactive diagrams, and includes flashcards for self-testing. Gemini first cited the sources in the notebook, then spent several minutes assembling the page.
Who Will Benefit and What the Caveats Are?
The most obvious scenario is studying. A student uploads lectures and a textbook, gets an interactive study guide, and shares the link with classmates. The second scenario is closer to work teams: from a set of internal documents, you can quickly assemble an understandable introductory primer for newcomers. The third is product and research notes, which are easier to show colleagues as a navigable structure than as a long block of text. There are also limitations worth keeping in mind.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.