notebooklm-py
Unofficial Python API, CLI and agent skill for Google Gemini Notebook (NotebookLM): sources, cited answers, and generation of podcasts, videos and reports
High risk
We rate an entry high when the tool writes to external systems, handles money, production databases or secrets, or runs arbitrary commands. The CLI installs it only with your consent.
Why this level
- Authenticates to a Google account and works through undocumented APIs, creating and deleting notebooks
- Master-token mode is a full-account credential that must be stored carefully
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/notebooklm-pyDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Install the tool
pipx install notebooklm-pyWithout pipx, pip install notebooklm-py works too.
Install the CLI: uv tool install "notebooklm-py[browser]", then notebooklm login to sign in to Google. Install the agent skill with notebooklm skill install or npx skills add teng-lin/notebooklm-py. Verify access with notebooklm auth check --test --json.
Other ways from the author
uv tool install "notebooklm-py[browser]"
notebooklm loginInstalls the CLI into an isolated environment and signs in to Google. The first sign-in auto-downloads Chromium, about 170 MB. You can use pipx instead of uv tool.
This is third-party code. Review the repository files before installing.
What it does
The project gives programmatic access to Gemini Notebook, formerly NotebookLM, through a Python API, the notebooklm CLI and a SKILL.md agent skill. From code or the command line you can create notebooks, add sources (links, PDF, YouTube, Google Drive, text), ask questions with source-grounded, cited answers, and run web research and deep analysis with auto-import of what is found. Separately it generates audio overviews (podcasts), videos, slide decks, infographics, quizzes, flashcards, reports, data tables and mind maps, all downloadable locally as MP3, MP4, PDF, PNG, CSV, JSON and Markdown, several of which the web UI does not offer. The skill teaches an agent such as Claude Code or Codex to work through the same CLI, and an optional MCP server in preview connects NotebookLM to MCP clients, including as a remote connector through a tunnel.
Who it is for. For developers, analysts and writers who need automatable access to NotebookLM for research and content generation.
Good fit when
- You need to bulk-add sources to a notebook and get cited answers from a script or agent
- You need to generate podcasts, videos, reports or quizzes from your material and download them in batches
- You need to offload heavy analysis of a document corpus to NotebookLM and only orchestrate with an agent
- You need to connect NotebookLM to Claude Code or Codex as a tool
Not a fit when
- A one-off task that is simpler to do by hand in the NotebookLM web UI
- You need stability guarantees: the library relies on undocumented Google APIs and can break
Example request
Create a notebook, add these PDFs to it and make a podcast overview from themLimitations
This is an unofficial library, not affiliated with Google; it uses undocumented APIs that can change at any time, and heavy use may hit rate limits. You need a Google account with NotebookLM access and a browser sign-in; for servers and headless use there is a durable master-token mode, which is a full-account credential, so the author advises keeping it on a dedicated account. NotebookLM itself is a Google service and is not reachable from Russia without a VPN. The CLI needs Python 3.10 or newer, and sign-in auto-downloads Chromium, about 170 MB. The MCP server is in preview and its tool surface may change between releases.
How to disable. Remove the CLI with uv tool uninstall notebooklm-py or pipx uninstall notebooklm-py, or drop the package from the project environment. Remove the installed skill from ~/.claude/skills/notebooklm and ~/.agents/skills/notebooklm. Auth files and profiles live in ~/.notebooklm.
Security check
- Authenticates to a Google account and works through undocumented APIs, creating and deleting notebooks
- Master-token mode is a full-account credential that must be stored carefully
README in short
The README describes notebooklm-py as a comprehensive skill and unofficial Python API for Gemini Notebook, formerly NotebookLM, with full programmatic access to its features via Python, CLI, MCP and agents. It lists patterns: offloading heavy analysis to NotebookLM to save tokens, cross-session memory for agents, cited answers over your own sources, generating and bulk-downloading material, scheduled runs and remote access from a phone. The CLI installs via uv tool or pipx followed by notebooklm login, the library installs from PyPI, and auth options include browser login, cookie import and a durable master token. An optional Android gRPC backend and a preview MCP server are documented separately. MIT licensed.
FAQ
Is this an official Google tool?
No, the project is unofficial and not affiliated with Google; it works through undocumented APIs that can change.
How do I sign in without a browser on a server?
Use master-token mode: notebooklm login --master-token --account you@example.com, then move master_token.json into the profile. It is a full-account credential, so keep it on a dedicated account.
Related
A self-hosted knowledge base with block-level references and a built-in MCP server for connecting AI agents to your notes
A CLI for every Google Workspace API with JSON output and agent skills: Drive, Gmail, Calendar, Sheets and more
Local search over Markdown notes, docs and meeting transcripts: keywords, semantic search and reranking, with an MCP server
A task manager for AI-driven development: breaks a PRD into dependent tasks and guides the agent through them via MCP or CLI