NLM - NotebookLM CLI

A powerful command-line interface for Google NotebookLM

⚠️ Important Disclaimer: This CLI uses internal APIs that are undocumented and may change without notice. Not affiliated with or endorsed by Google. Use at your own risk for personal/experimental purposes. See also: notebooklm-mcp for MCP server integration.

🎬 Demo

Watch a ~12 minute overview of the CLI in action:

✨ Features

• Full NotebookLM API Coverage — Notebooks, sources, audio podcasts, reports, quizzes, flashcards, mind maps, slides, infographics, videos, and data tables
• Seamless Authentication — Uses Chrome DevTools Protocol for reliable, automatic cookie extraction
• AI-Teachable — Run nlm --ai to output comprehensive documentation that AI assistants can consume
• Alias System — Create memorable shortcuts for long UUIDs (e.g., myproject instead of abc123-def456-...)
• Multiple Output Formats — Rich tables, JSON, quiet (IDs only), or full details
• Profile Support — Manage multiple Google accounts with named profiles
• Research Integration — Deep web search or Google Drive search to discover and import sources

📦 Installation

Install from PyPI using your preferred package manager:

# Using pip
pip install notebooklm-cli

# Using pipx (recommended for CLI tools)
pipx install notebooklm-cli

# Using uv
uv tool install notebooklm-cli

Requirements:

• Python 3.10+
• Google Chrome (for authentication)

🚀 Quick Start

1. Authenticate

nlm login

This launches Chrome, navigates to NotebookLM, and automatically extracts your session cookies. You'll need to log in to your Google account if not already signed in.

2. List Your Notebooks

nlm notebook list

3. Create a Notebook and Add Sources

# Create a new notebook
nlm notebook create "My Research"
# Output: Created notebook: abc123-def456-...

# Add a URL source
nlm source add abc123-def456 --url "https://example.com/article"

# Add a YouTube video
nlm source add abc123-def456 --url "https://youtube.com/watch?v=..."

# Add pasted text
nlm source add abc123-def456 --text "Your content here" --title "My Notes"

4. Generate a Podcast

nlm audio create abc123-def456 --confirm

5. Check Generation Status

nlm studio status abc123-def456

🏷️ Aliases (UUID Shortcuts)

Tired of typing long UUIDs? Create aliases:

# Set an alias
nlm alias set myproject abc123-def456-... # Types are auto-detected!

# Now use the alias anywhere
nlm notebook get myproject
nlm source list myproject
nlm audio create myproject --confirm

# Manage aliases
nlm alias list              # List all aliases
nlm alias get myproject     # Resolve to UUID
nlm alias delete myproject  # Remove alias

🤖 AI Integration

Option 1: Quick Context (nlm --ai)

The --ai flag outputs comprehensive, structured documentation designed for AI assistants:

nlm --ai

This prints a 400+ line guide covering all commands with exact syntax, authentication flow, error handling, complete task sequences, and 12 tips for AI automation.

Use case: Paste the output of nlm --ai into your AI assistant's context to teach it how to use the CLI.

Option 2: AI Skill (nlm-cli-skill/)

For AI coding assistants that support skills (Claude Code, Gemini CLI/Antigravity, OpenAI Codex, etc.), this repo includes a comprehensive skill:

nlm-cli-skill/
├── SKILL.md              # Main skill file with 10 critical rules
└── references/
    ├── command_reference.md   # Complete command signatures
    ├── troubleshooting.md     # Error diagnosis & solutions
    └── workflows.md           # End-to-end task sequences

To use: Copy the nlm-cli-skill/ folder to your AI tool's skills directory, or point your tool to this folder directly.

📚 Command Reference

Core Commands

Source Management

Research (Discover New Sources)

Content Generation

All generation commands require --confirm (or -y) to execute:

Studio (Artifact Management)

Chat (Interactive Q&A)

Chat REPL commands: /sources, /clear, /help, /exit

Configuration

Authentication

🎛️ Output Formats

Most list commands support multiple output formats:

nlm notebook list              # Rich table (default)
nlm notebook list --json       # JSON output
nlm notebook list --quiet      # IDs only (for scripting)
nlm notebook list --title      # "ID: Title" format
nlm source list --url          # "ID: URL" format
nlm notebook list --full       # All columns

👤 Profiles (Multiple Accounts)

Manage multiple Google accounts with named profiles:

# Login to a specific profile
nlm login --profile work
nlm login --profile personal

# Use a profile for commands
nlm notebook list --profile work

# List all profiles
nlm auth list

# Delete a profile
nlm auth delete work --confirm

⌨️ Shell Completion

Enable tab completion for faster command entry:

# Auto-install for your current shell
nlm --install-completion

# Or show the completion script to install manually
nlm --show-completion

⚠️ Session Lifetime

NotebookLM sessions typically last ~20 minutes. If commands start failing with authentication errors, simply re-run:

nlm login

🔧 Troubleshooting

Having issues? See the Troubleshooting Guide for solutions to common problems including authentication, network issues, and OpenAI Codex sandbox configuration.

📖 Documentation

For detailed technical documentation on the internal API and advanced usage, see the docs/ folder:

• Troubleshooting — Common issues and solutions
• CLI Test Plan — End-to-end testing procedures
• Technical Deep Dive — Internal API details

For AI assistants, run nlm --ai to get the full command reference.

🤝 Contributing

Contributions are welcome! See CONTRIBUTING.md for guidelines.

# Quick start for contributors
git clone https://github.com/jacob-bd/notebooklm-cli.git
cd notebooklm-cli
uv pip install -e ".[dev]"
uv run pytest

⚠️ Limitations

• Rate limits: Free tier has ~50 queries/day
• No official support: API may change without notice
• Cookie expiration: Need to re-authenticate every few weeks

🎨 Vibe Coding Alert

Full transparency: this project was built by a non-developer using AI coding assistants. If you're an experienced Python developer, you might look at this codebase and wince. That's okay.

The goal here was to scratch an itch—programmatic access to NotebookLM—and learn along the way. The code works, but it's likely missing patterns, optimizations, or elegance that only years of experience can provide.

This is where you come in. If you see something that makes you cringe, please consider contributing rather than just closing the tab. PRs and issues are welcome.

📄 License

MIT License. See LICENSE for details.