PyNotebookLM

Production-grade Python library for Google NotebookLM automation.

Documentation: https://github.com/Experto-AI/pynotebooklm

Features

• 🔐 Secure Authentication - Browser-based Google login with cookie persistence
• 📓 Notebook Management - Create, list, rename, and delete notebooks
• 📰 Source Management - Add URLs, YouTube videos, Google Drive docs, and text
• 🔍 Research & Analysis - Query notebooks and discover related sources (Web/Google Drive)
• 🧠 Mind Maps - Generate, save, list, and export mind maps (JSON/OPML/FreeMind)
• 🎙️ Content Generation - Create audio overviews (podcasts), videos, infographics, and slides
• 📚 Study Tools - Create flashcards, quizzes, and briefing documents

Installation

pip install pynotebooklm

Or with Poetry:

poetry add pynotebooklm

Install Playwright Browsers

After installing, you need to install the Playwright browser:

playwright install chromium

Quick Start

1. Authenticate

First, login to NotebookLM with your Google account:

pynotebooklm auth login

This opens a browser window for you to login. Cookies are saved to ~/.pynotebooklm/auth.json.

2. Verify Authentication

pynotebooklm auth check

3. Use the Library

import asyncio
from pynotebooklm import NotebookLMClient

async def main():
    async with NotebookLMClient() as client:
        # List notebooks
        notebooks = await client.notebooks.list()
        for nb in notebooks:
            print(f"{nb.name} ({nb.id})")

asyncio.run(main())

CLI Commands

Authentication

pynotebooklm auth login              # Login with Google account (opens browser)
pynotebooklm auth check              # Check authentication status
pynotebooklm auth logout             # Clear saved authentication

Notebooks

pynotebooklm notebooks list                        # List all notebooks
pynotebooklm notebooks list --detailed             # With source count and dates
pynotebooklm notebooks create "My Notebook"        # Create a new notebook
pynotebooklm notebooks get <notebook_id>           # Get detailed notebook info with sources
pynotebooklm notebooks describe <notebook_id>      # Get AI-generated summary and topics
pynotebooklm notebooks rename <notebook_id> "New Name"  # Rename a notebook
pynotebooklm notebooks delete <notebook_id>        # Delete a notebook
pynotebooklm notebooks delete <notebook_id> -f     # Delete without confirmation

Sources

pynotebooklm sources add <notebook_id> <url>               # Add URL source (Web or YouTube)
pynotebooklm sources add-text <notebook_id> "text"         # Add plain text source
pynotebooklm sources add-drive <notebook_id> <drive_id>    # Add Google Drive source
pynotebooklm sources list <notebook_id>                    # List sources in notebook
pynotebooklm sources list <notebook_id> --check-freshness  # Check if Drive sources are up-to-date
pynotebooklm sources describe <source_id>                  # AI-generated summary and keywords
pynotebooklm sources get-text <source_id>                  # Extract raw indexed text
pynotebooklm sources sync <source_id>                      # Sync stale Drive source
pynotebooklm sources delete <notebook_id> <source_id>      # Delete a source
pynotebooklm sources delete <notebook_id> <source_id> -f   # Delete without confirmation

Research Discovery

pynotebooklm research start <notebook_id> "topic"              # Fast web research
pynotebooklm research start <notebook_id> "topic" --deep       # Deep research (more comprehensive)
pynotebooklm research start <notebook_id> "topic" --source drive   # Search Google Drive
pynotebooklm research poll <notebook_id>                       # Check status and get results
pynotebooklm research poll <notebook_id> --auto-import         # Poll & auto-import when done
pynotebooklm research import <notebook_id>                     # Import all discovered sources
pynotebooklm research import <notebook_id> --indices 0,1,2     # Import specific sources

Mind Maps

pynotebooklm mindmap create <notebook_id>                      # Create from all sources
pynotebooklm mindmap create <notebook_id> --title "My Map"     # With custom title
pynotebooklm mindmap list <notebook_id>                        # List existing maps
pynotebooklm mindmap export <notebook_id> <map_id> -f json     # Export to JSON
pynotebooklm mindmap export <notebook_id> <map_id> -f opml     # Export to OPML
pynotebooklm mindmap export <notebook_id> <map_id> -f freemind # Export to FreeMind (.mm)

Chat & Query

pynotebooklm query ask <notebook_id> "question"                # Ask a question
pynotebooklm query ask <notebook_id> "question" --sources <ids> # Scope to specific sources
pynotebooklm query ask <notebook_id> "question" -c <id>        # Follow-up in conversation
pynotebooklm query summary <notebook_id>                       # Get AI summary
pynotebooklm query briefing <notebook_id>                      # Create briefing document
pynotebooklm query configure <notebook_id> --goal learning     # Set conversation goal
pynotebooklm query configure <notebook_id> --length longer     # Set response length

Content Generation

# Audio Overview (Podcast)
pynotebooklm generate audio <notebook_id>                      # Generate with defaults
pynotebooklm generate audio <notebook_id> --format deep_dive   # Format: deep_dive, brief, critique, debate
pynotebooklm generate audio <notebook_id> --length short       # Length: short, default, long
pynotebooklm generate audio <notebook_id> --language es        # Language: en, es, fr, de, ja, etc.
pynotebooklm generate audio <notebook_id> --focus "key topics" # Focus prompt for AI

# Video Overview
pynotebooklm generate video <notebook_id>                      # Generate with defaults
pynotebooklm generate video <notebook_id> --format explainer   # Format: explainer, brief
pynotebooklm generate video <notebook_id> --style anime        # Style: auto_select, classic, whiteboard, kawaii, anime, watercolor, retro_print, heritage, paper_craft

# Infographic
pynotebooklm generate infographic <notebook_id>                    # Generate with defaults
pynotebooklm generate infographic <notebook_id> --orientation portrait  # Orientation: landscape, portrait, square
pynotebooklm generate infographic <notebook_id> --detail detailed  # Detail: concise, standard, detailed

# Slide Deck
pynotebooklm generate slides <notebook_id>                         # Generate with defaults
pynotebooklm generate slides <notebook_id> --format presenter_slides  # Format: detailed_deck, presenter_slides
pynotebooklm generate slides <notebook_id> --length short          # Length: short, default

### Study Tools

```bash
# Flashcards
pynotebooklm study flashcards <notebook_id>                    # Generate flashcards
pynotebooklm study flashcards <notebook_id> --difficulty hard  # Difficulty: easy, medium, hard

# Quiz
pynotebooklm study quiz <notebook_id>                          # Generate quiz (default 2 questions)
pynotebooklm study quiz <notebook_id> --questions 10           # Number of questions
pynotebooklm study quiz <notebook_id> --difficulty 3           # Difficulty: 1 (easy) to 3 (hard)

# Data Table
pynotebooklm study table <notebook_id> --description "Extract dates"     # Extract structured data
pynotebooklm study table <notebook_id> --description "..." --language es # Output language

pynotebooklm studio status <notebook_id> # shows study artifacts

Studio (Artifact Management)

pynotebooklm studio list <notebook_id>             # List all artifacts (Briefings, Audio, Video, etc.)
pynotebooklm studio status <notebook_id>           # Detailed status with download URLs
pynotebooklm studio delete <artifact_id>           # Delete an artifact (with confirmation)
pynotebooklm studio delete <artifact_id> --force   # Delete without confirmation

Troubleshooting

Authentication Errors

• Run pynotebooklm auth login to refresh cookies.
• Delete ~/.pynotebooklm/auth.json if cookies are corrupted, then login again.

Playwright Issues

• Ensure the browser is installed: playwright install chromium
• On Linux, install required system deps (see Dockerfile for a reference list).

Debugging Requests

• Enable verbose logging with PYNOTEBOOKLM_DEBUG=1.
• Enable structured telemetry with PYNOTEBOOKLM_TELEMETRY=1.

Development

Setup

# Clone the repository
git clone https://github.com/Experto-AI/pynotebooklm.git
cd pynotebooklm

# Install dependencies
poetry install

# Install Playwright browsers
poetry run playwright install chromium

Running Tests

# Run all unit tests
poetry run pytest tests/unit/ -v

# Run with coverage
poetry run pytest tests/unit/ -v --cov=src/pynotebooklm

# Run specific test file
poetry run pytest tests/unit/test_auth.py -v

Linting and Formatting

# Lint with ruff
poetry run ruff check src tests

# Format with black
poetry run black src tests

# Type check with mypy
poetry run mypy src

Project Structure

pynotebooklm/
├── src/pynotebooklm/
│   ├── __init__.py        # Public API exports
│   ├── auth.py            # Authentication manager
│   ├── session.py         # Browser session management
│   ├── api.py             # Low-level RPC wrapper
│   ├── notebooks.py       # Notebook management
│   ├── sources.py         # Source management
│   ├── research.py        # Research discovery
│   ├── mindmaps.py        # Mind map generation
│   ├── study.py           # Study tools (flashcards, quiz, tables)
│   ├── content.py         # Content generation (audio, video, etc.)
│   ├── chat.py            # Chat and query functionality
│   ├── models.py          # Pydantic data models
│   ├── exceptions.py      # Custom exceptions
│   └── cli.py             # CLI interface
├── tests/
│   ├── unit/              # Unit tests (457)
│   ├── integration/       # Integration tests (192)
│   └── fixtures/          # Mock responses
└── docs/                  # Documentation

How It Works

PyNotebookLM uses browser automation (Playwright) to interact with NotebookLM's internal APIs:

1. Authentication: Opens a browser for Google login, extracts cookies
2. Session: Creates a headless browser session with injected cookies
3. API Calls: Executes fetch() requests via page.evaluate() in browser context
4. RPC Protocol: Communicates using NotebookLM's internal RPC format

This approach provides:

• ✅ Full feature access (all 31 tools)
• ✅ No API keys needed
• ✅ Works with consumer Google accounts
• ⚠️ Requires browser automation
• ⚠️ Cookie refresh needed every 2-4 weeks

License

Apache License 2.0 - see LICENSE for details.

Disclaimer

This is an unofficial library. It uses NotebookLM's internal APIs which may change without notice. Use at your own risk.