Documentation Index
Fetch the complete documentation index at: https://mintlify.com/teng-lin/notebooklm-py/llms.txt
Use this file to discover all available pages before exploring further.
The notebooklm CLI exposes every feature of the Python library as shell commands organized into logical groups. You can run it interactively, embed it in shell scripts, or call it from CI/CD pipelines. Most commands accept a --json flag for machine-readable output and support partial notebook IDs for convenience.
Command structure
Every invocation follows this shape:
notebooklm [-p PROFILE] [--storage PATH] [--version] [-v] <command> [OPTIONS] [ARGS]
Global options
| Option | Description |
|---|
-p, --profile NAME | Use a named profile (overrides NOTEBOOKLM_PROFILE) |
--storage PATH | Override the default storage location |
-v, --verbose | Enable verbose output |
--version | Show version and exit |
--help | Show help message |
Environment variables
| Variable | Description |
|---|
NOTEBOOKLM_HOME | Base directory for all config files (default: ~/.notebooklm) |
NOTEBOOKLM_PROFILE | Active profile name (default: default) |
NOTEBOOKLM_AUTH_JSON | Inline authentication JSON (for CI/CD, no file writes needed) |
NOTEBOOKLM_DEBUG_RPC | Enable RPC debug logging—set to 1 to activate |
Session commands
Session commands manage authentication and the active notebook context. You must authenticate before using any other command.
| Command | Description | Example |
|---|
login | Authenticate via browser | notebooklm login |
login --browser msedge | Use Edge for SSO-gated orgs | notebooklm login --browser msedge |
use <id> | Set active notebook | notebooklm use abc123 |
status | Show current context | notebooklm status |
status --paths | Show config file paths | notebooklm status --paths |
status --json | JSON output | notebooklm status --json |
clear | Clear active notebook | notebooklm clear |
auth check | Diagnose auth issues | notebooklm auth check |
auth check --test | Validate with a live network test | notebooklm auth check --test |
doctor | Check environment health | notebooklm doctor |
doctor --fix | Auto-fix detected issues | notebooklm doctor --fix |
Partial ID matching
notebooklm use <id> and most ID-accepting commands support partial prefix matching. Providing the first six or more characters of a UUID is enough, as long as the prefix is unique.
notebooklm use abc # Matches any notebook ID starting with "abc"
Profile commands
Named profiles let you manage multiple Google accounts. Each profile has its own session file and active notebook context.
| Command | Description | Example |
|---|
profile list | List all profiles | notebooklm profile list |
profile create <name> | Create a profile | notebooklm profile create work |
profile switch <name> | Set the active profile | notebooklm profile switch work |
profile delete <name> | Delete a profile | notebooklm profile delete old |
profile rename <old> <new> | Rename a profile | notebooklm profile rename old new |
Notebook commands
| Command | Description | Example |
|---|
list | List all notebooks | notebooklm list |
create <title> | Create a notebook | notebooklm create "Research" |
delete <id> | Delete a notebook | notebooklm delete abc123 |
rename <title> | Rename the current notebook | notebooklm rename "New Title" |
summary | Get the AI-generated summary | notebooklm summary |
Chat commands
The ask command sends a question to the active notebook and prints the answer. Use --save-as-note to persist the response as a notebook note.
| Command | Description | Example |
|---|
ask <question> | Ask a question | notebooklm ask "What is this about?" |
ask -s <id> | Limit to specific sources | notebooklm ask "Summarize" -s src1 -s src2 |
ask --json | Include source references | notebooklm ask "Explain X" --json |
ask --save-as-note | Save response as a note | notebooklm ask "Explain X" --save-as-note |
ask --save-as-note --note-title "T" | Save with a custom title | notebooklm ask "Explain X" --save-as-note --note-title "T" |
configure --mode learning-guide | Set the chat persona | notebooklm configure --mode learning-guide |
history | View conversation history | notebooklm history |
history --clear | Clear local conversation cache | notebooklm history --clear |
history --save | Save history as a note | notebooklm history --save |
Source commands
source add auto-detects the content type from its argument: HTTP URLs become web sources, YouTube URLs extract transcripts, and file paths trigger uploads.
| Command | Arguments | Options | Example |
|---|
source list | — | — | notebooklm source list |
source add <content> | URL / file path / text | — | notebooklm source add "https://..." |
source add-drive <id> <title> | Drive file ID, title | — | notebooklm source add-drive abc123 "Doc" |
source add-research <query> | Search query | --mode [fast|deep], --from [web|drive], --import-all, --no-wait | notebooklm source add-research "AI" --mode deep --no-wait |
source get <id> | Source ID | — | notebooklm source get src123 |
source fulltext <id> | Source ID | --json, -o FILE | notebooklm source fulltext src123 -o content.txt |
source guide <id> | Source ID | --json | notebooklm source guide src123 |
source rename <id> <title> | Source ID, new title | — | notebooklm source rename src123 "New Name" |
source refresh <id> | Source ID | — | notebooklm source refresh src123 |
source delete <id> | Source ID | — | notebooklm source delete src123 |
source wait <id> | Source ID | --timeout, --interval | notebooklm source wait src123 |
Research commands
Use research commands to check on and wait for background web or Drive research started with source add-research --no-wait.
| Command | Options | Example |
|---|
research status | --json | notebooklm research status |
research wait | --timeout, --interval, --import-all, --json | notebooklm research wait --import-all |
Generate commands
All generate subcommands support --source/-s (repeatable) to target specific sources, --language for multilingual output, --json for machine-readable task IDs, and --retry N for automatic exponential backoff on rate limits. Most are asynchronous by default—use --wait to block until completion.
| Command | Options | Example |
|---|
generate audio [desc] | --format [deep-dive|brief|critique|debate], --length [short|default|long], --wait | notebooklm generate audio "Focus on history" --wait |
generate video [desc] | --format [explainer|brief|cinematic], --style [auto|classic|whiteboard|kawaii|anime|watercolor|retro-print|heritage|paper-craft], --wait | notebooklm generate video --style whiteboard --wait |
generate cinematic-video [desc] | Alias for video --format cinematic; same options | notebooklm generate cinematic-video "Documentary summary" --wait |
generate slide-deck [desc] | --format [detailed|presenter], --length [default|short], --wait | notebooklm generate slide-deck --wait |
generate revise-slide <desc> | -a/--artifact <id> (required), --slide N (required), --wait | notebooklm generate revise-slide "Move title up" --artifact <id> --slide 0 |
generate quiz [desc] | --difficulty [easy|medium|hard], --quantity [fewer|standard|more], --wait | notebooklm generate quiz --difficulty hard --wait |
generate flashcards [desc] | --difficulty [easy|medium|hard], --quantity [fewer|standard|more], --wait | notebooklm generate flashcards --wait |
generate infographic [desc] | --orientation [landscape|portrait|square], --detail [concise|standard|detailed], --wait | notebooklm generate infographic --orientation portrait --wait |
generate report [desc] | --format [briefing-doc|study-guide|blog-post|custom], --append "...", --wait | notebooklm generate report --format study-guide --wait |
generate data-table <desc> | --wait | notebooklm generate data-table "compare key concepts" --wait |
generate mind-map | (synchronous, no --wait option) | notebooklm generate mind-map |
Artifact commands
| Command | Arguments | Options | Example |
|---|
artifact list | — | --type | notebooklm artifact list --type audio |
artifact get <id> | Artifact ID | — | notebooklm artifact get art123 |
artifact rename <id> <title> | Artifact ID, title | — | notebooklm artifact rename art123 "Title" |
artifact delete <id> | Artifact ID | — | notebooklm artifact delete art123 |
artifact export <id> | Artifact ID | --type [docs|sheets], --title | notebooklm artifact export art123 --type sheets |
artifact poll <task_id> | Task ID | — | notebooklm artifact poll task123 |
artifact wait <id> | Artifact ID | --timeout, --interval | notebooklm artifact wait art123 |
Download commands
Download commands write files to your local machine. Use --all for batch downloads, --latest for the most recent artifact, or -a/--artifact to target a specific ID.
| Command | Options | Default format | Example |
|---|
download audio [path] | --all, --latest, -a, --force | .mp4 | notebooklm download audio ./podcast.mp3 |
download video [path] | --all, --latest, -a, --force | .mp4 | notebooklm download video ./overview.mp4 |
download cinematic-video [path] | Same as video | .mp4 | notebooklm download cinematic-video ./documentary.mp4 |
download slide-deck [path] | --format [pdf|pptx], --all, --latest | .pdf | notebooklm download slide-deck --format pptx |
download infographic [path] | --all, --latest, -a | .png | notebooklm download infographic ./info.png |
download report [path] | --all, --latest, -a | .md | notebooklm download report ./study-guide.md |
download mind-map [path] | --all, --latest, -a | .json | notebooklm download mind-map ./map.json |
download data-table [path] | --all, --latest, -a | .csv | notebooklm download data-table ./data.csv |
download quiz [path] | --format [json|markdown|html], -a | .json | notebooklm download quiz --format markdown quiz.md |
download flashcards [path] | --format [json|markdown|html], -a | .json | notebooklm download flashcards cards.json |
Language commands
Language is a global setting that applies to all artifact generation in your account.
| Command | Description | Example |
|---|
language list | List all 80+ supported languages | notebooklm language list |
language get | Show current language setting | notebooklm language get |
language get --local | Skip server sync | notebooklm language get --local |
language set <code> | Set the output language | notebooklm language set zh_Hans |
language set <code> --local | Save locally only | notebooklm language set ja --local |
Sharing commands
| Command | Description | Example |
|---|
share status | Show sharing configuration | notebooklm share status |
share public --enable | Enable public link sharing | notebooklm share public --enable |
share public --disable | Disable public sharing | notebooklm share public --disable |
share view-level full | Set viewers to see full notebook | notebooklm share view-level full |
share view-level chat | Restrict viewers to chat only | notebooklm share view-level chat |
share add <email> | Share with a user (viewer by default) | notebooklm share add user@example.com |
share update <email> | Change a user’s permission | notebooklm share update user@example.com --permission editor |
share remove <email> | Remove user access | notebooklm share remove user@example.com |
Note commands
| Command | Arguments | Example |
|---|
note list | — | notebooklm note list |
note create <content> | Note text | notebooklm note create "My notes..." |
note get <id> | Note ID | notebooklm note get note123 |
note save <id> | Note ID + --title, --content | notebooklm note save note123 --title "Updated title" |
note rename <id> <title> | Note ID, title | notebooklm note rename note123 "Title" |
note delete <id> | Note ID | notebooklm note delete note123 |
Skill commands
Skill commands install the NotebookLM skill into local agent directories so tools like Claude Code can pick it up automatically.
| Command | Description | Example |
|---|
skill install | Install to ~/.claude/skills/ and ~/.agents/skills/ | notebooklm skill install |
skill status | Check installed targets and version | notebooklm skill status |
skill uninstall | Remove installed targets | notebooklm skill uninstall |
skill show | Display the packaged skill file | notebooklm skill show |
Common workflows
These end-to-end examples show how to combine CLI commands into useful pipelines.
Research → podcast
notebooklm create "Climate Change Research"
notebooklm use <notebook_id>
notebooklm source add "https://en.wikipedia.org/wiki/Climate_change"
notebooklm source add-research "climate change policy 2024" --mode deep --import-all
notebooklm generate audio "Focus on policy solutions" --format debate --wait
notebooklm download audio ./climate-podcast.mp3
Document analysis → study materials
notebooklm create "Exam Prep"
notebooklm use <id>
notebooklm source add "./textbook-chapter.pdf"
notebooklm source add "./lecture-notes.pdf"
notebooklm summary
notebooklm generate quiz --difficulty hard --wait
notebooklm generate flashcards --wait
notebooklm generate report --format study-guide --wait
notebooklm download quiz --format markdown quiz.md
notebooklm download flashcards cards.json
notebooklm download report ./study-guide.md
YouTube → quick summary
notebooklm create "Video Notes"
notebooklm use <id>
notebooklm source add "https://www.youtube.com/watch?v=VIDEO_ID"
notebooklm summary
notebooklm ask "What are the main points?"
notebooklm generate report --format briefing-doc --wait
notebooklm download report ./briefing.md
Bulk import
notebooklm use <id>
notebooklm source add "https://example.com/article1"
notebooklm source add "https://example.com/article2"
# Loop over local files
for f in ./papers/*.pdf; do
notebooklm source add "$f"
done
Use --wait when you need a file on disk immediately. For long-running generation (audio, video, slide decks), omit --wait and check back with notebooklm artifact list, or use artifact wait <id> in a background task.Nearly every command supports --json for structured output, making it easy to parse results with jq or pass IDs between steps in a script.
Tips for LLM agents
When this CLI is called programmatically by an AI agent, keep the following in mind.
- Use
notebooklm use <id> to set context for a single-agent session, or pass -n <id> directly to commands in parallel workflows to avoid context collisions.
- Generation commands are async by default except
mind-map, which is synchronous. Use --no-wait for long operations and poll with artifact wait in a subagent.
- Partial IDs work for most commands—six or more characters is usually enough.
- Use
--json on every command to get machine-readable output instead of styled tables.
source add auto-detects the content type from the argument, so you do not need separate commands for URLs, files, or YouTube links.- For non-blocking deep research, run
source add-research --mode deep --no-wait and then research wait --import-all in a subagent.
