Provides tools for managing and interacting with Google NotebookLM, including listing and creating notebooks, managing sources, and querying the NotebookLM AI with conversation history support.
Allows for the management of Google Drive documents as sources within NotebookLM notebooks, enabling the AI to analyze and answer questions based on Drive-hosted content.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@NotebookLM MCP Serversummarize the main insights from my 'Project Phoenix' notebook"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
NotebookLM MCP Server
This repository provides a Model Context Protocol (MCP) server for NotebookLM. It allows AI assistants (like Claude, Antigravity, or others supporting MCP) to interact with your NotebookLM notebooks, sources, and conversations.
Features
List Notebooks: View all your NotebookLM notebooks.
Create Notebooks: Programmatically create new notebooks.
Manage Sources: Add websites, Google Drive documents, or pasted text to your notebooks.
Query Notebooks: Ask questions about your sources using the NotebookLM AI.
Conversation History: Full support for follow-up questions and conversation context.
Auto-Save Notes: Automatically save AI responses as notes in your notebooks.
Prerequisites
Python 3.10 or higher.
A Google Account with access to NotebookLM.
An MCP-compatible client (e.g., Cursor, Claude Desktop).
Quick Start
1. Clone the repository
2. Set up the environment
Install dependencies:
Authenticate with NotebookLM:
Follow the prompts to authorize the application. This will create a local auth.json file in ~/.notebooklm-mcp/.
3. Configure your MCP Client
For Cursor:
Copy
mcp_config.json.exampleto your Cursor config directory:Linux:
~/.config/cursor/mcp.jsonmacOS:
~/Library/Application Support/Cursor/mcp.jsonWindows:
%APPDATA%\Cursor\mcp.json
Edit the
commandfield with the absolute path tonotebooklm-mcp:{ "mcpServers": { "notebooklm": { "command": "/home/YOUR_USER/.local/bin/notebooklm-mcp", "args": [], "env": {} } } }Find the binary path:
which notebooklm-mcp # or ls ~/.local/bin/notebooklm-mcpRestart Cursor to apply the configuration.
Usage Examples
Basic Usage
Test your setup by listing notebooks:
Or query a notebook directly:
Auto-Save Notes Feature
The repository includes an automatic note-saving feature that saves all AI responses as notes in your notebooks. This is especially useful when working through MCP API, as responses aren't automatically saved in the web interface history.
Quick start:
See docs/AUTO_SAVE_NOTES.md for detailed documentation (if available locally).
Security Note
Your authentication tokens are stored locally in~/.notebooklm-mcp/auth.json. Never share this file or commit it to a public repository. The .gitignore in this repo is already configured to ignore this folder.
License
MIT