Names, versions and owners change fast. Checked October 2026.
In plain words
NotebookLM is an app from Google, not a general chatbot. You make a notebook, add your own sources, such as PDFs, Word files, web pages, YouTube videos or photos, and then ask questions about them.
The answers are grounded in your sources. Each claim comes with a small numbered citation you can click to see the exact passage it came from. That makes it handy for studying from class notes, reading a long report or preparing for a meeting.
It can also turn your sources into other formats, such as written summaries, Video Overviews and Audio Overviews: a spoken walk-through of your material you can listen to like a short podcast. Under the hood it uses Google's Gemini models and a method called retrieval-augmented generation: find the relevant passages first, then write.
Everyday example
An open-book exam
In a closed-book exam you answer from memory, and sometimes you misremember. In an open-book exam you may only use the textbook on your desk, and you can point to the page for every answer. A general chatbot is the closed-book student. NotebookLM is the open-book student: it answers from the sources you put on its desk and shows the page.
Where the example breaks: An open-book student still understands the subject. NotebookLM can still misread a passage or join two facts wrongly, so the citations are there for you to check.
Latestas of October 2026
- Name
- Renamed Gemini Notebook on 16 July 2026; still a standalone app
- Where it appears
- Its own app and website, plus notebooks inside the Gemini app with full syncing; Google has said notebooks are coming to AI Mode in Search
- New in 2026
- Each notebook can use a secure cloud computer to write and run code for data analysis, first for top-tier and some business plans
These details change often. The rest of this page explains what it is, which stays true.
How it actually works
Every answer follows the look-up-then-write pattern, limited to your notebook.
Your sources are split and indexed
Each uploaded source is broken into passages and turned into embeddings, so the system can find passages that match a question by meaning, not just by exact words.
The relevant passages are retrieved
When you ask something, it picks the passages most likely to hold the answer and places them in the Gemini model's context window together with your question.
The model writes with citations
The model writes an answer from those passages and tags each claim with the source it used. Keeping answers tied to your material reduces, but does not remove, hallucination.
Same sources, new formats
The same retrieved material can be rewritten as a summary or a script, which text-to-speech then reads aloud as an Audio Overview, or turned into a short explainer video called a Video Overview.
- Made by
- Google, first as an experiment in Google Labs
- First available
- Early access from July 2023; opened to US users aged 18 and over on 8 December 2023
- Models inside
- Google's Gemini models
- Method
- Retrieval-augmented generation over sources you add
Because it answers from what you add, the quality of an answer depends on the quality of your sources.
Questions people ask
Can NotebookLM be wrong?
Yes. It can misread a passage, miss something important or join two facts incorrectly, and it is only as good as the sources you add. Because each claim is cited, you can click through and check the original passage.
Is NotebookLM called Gemini Notebook?
Google announced on 16 July 2026 that it was renaming NotebookLM to Gemini Notebook, keeping it as its own app, with notebooks also available inside the Gemini app. Many people still say NotebookLM; the box above has the current name.
How is NotebookLM different from ChatGPT or Gemini?
A general chatbot answers from everything it learned plus, sometimes, a web search. NotebookLM answers mainly from the sources you upload to a notebook and cites the passage behind each claim, which suits studying and research on specific material.
What is an Audio Overview?
A feature that turns the sources in a notebook into a spoken summary you can listen to, like a short podcast about your own material. Google says Audio Overviews work in more than 80 languages, in short and full-length versions.