notebooklm.md Specification#

Location: <material_directory>/notebooklm.md, for example, the file notebooklm.md in a directory like 20260301-How-to-make-openclaw-productive. This file is used to declare the source file manifest for NotebookLM projects and the types and parameters of planned artifact generation.

Below is a simple example that generates both audio and video files. See below for explanations of each parameter.

## global

language = "en"

## sources

- https://example.com/article

## audio

enabled = true
format = "DEEP_DIVE"
length = "DEFAULT"

## report

enabled = true
format = "BRIEFING_DOC"
append = "Add real-world use cases"

File Format#

notebooklm.md uses Markdown format and consists of the following sections:

1. Global configuration section - ## global block, sets global parameters (such as language)
2. Source declaration section - ## sources block, lists source files to add to NotebookLM notebook, including local files and web links. Note that links must be accessible without password or login.
3. Generation configuration sections - Multiple ## <artifact_type> blocks, declaring the artifacts to generate and their parameters

Global Configuration (global)#

Set global parameters in the ## global section:

Parameter Description:

• language - Language code for generated content

  • en - English ✓ Recommended for English projects

  • zh_Hans - Simplified Chinese

  • zh_Hant - Traditional Chinese

  • ja - Japanese

  • ko - Korean

  • es - Spanish

  • fr - French

  • de - German

Source Declaration (sources)#

Declare sources to add to notebook: automatically scan project files and/or remote URLs.

## sources

- https://example.com/article
- https://www.youtube.com/watch?v=...
- docs/architecture.md
- README.md

How It Works#

Auto-scan project files: All files in the project root directory (except hidden files, the outputs directory, and notebooklm.md) are automatically added to the notebook.

Remote URLs: Lines starting with http:// or https:// are recognized as URLs, supporting public web resources like articles and YouTube videos.

Local file paths: Local file paths relative to the project root directory, supporting multiple formats including Markdown, PDF, code files, audio/video, and images.

Artifact Generation Configuration#

Audio (Audio)#

Generate NotebookLM-style AI host podcast audio.

Declare in the ## audio section:

Parameter Details:

Video (Video Explanation)#

Generate animated video explanations.

Declare in the ## video section:

Parameter Details:

Cinematic Video (Cinematic Video)#

Advanced video generation requiring Google AI Ultra subscription.

Declare in the ## cinematic_video section:

Report (Report Document)#

Generate structured text reports.

Declare in the ## report section:

Parameter Details:

Quiz (Quiz)#

Generate multiple-choice or short-answer quizzes.

Declare in the ## quiz section:

Parameter Details:

Flashcards (Study Cards)#

Generate Anki-compatible study cards.

Declare in the ## flashcards section:

Parameter Details:

Infographic (Infographic)#

Generate visual information graphics.

Declare in the ## infographic section:

Parameter Details:

Slide Deck (Slide Deck)#

Generate presentation slides downloadable as PDF or PPTX.

Declare in the ## slide_deck section:

Parameter Details:

Data Table (Data Table)#

Generate structured data tables (CSV format).

Declare in the ## data_table section:

Parameter Description:

• enabled - Whether to generate
• description - Describe table topic and comparison dimensions

Mind Map (Mind Map)#

Generate hierarchical mind maps (JSON format).

Declare in the ## mind_map section:

Note: Mind Map is generated synchronously with no additional parameters.

Complete Examples#

Minimal Configuration (Audio + Report only)#

## global

language = "en"

## sources

- https://example.com/article

## audio

enabled = true
format = "DEEP_DIVE"
length = "DEFAULT"

## report

enabled = true
format = "BRIEFING_DOC"
append = "Add real-world use cases"

Complete Configuration (Multiple artifact types)#

## global

language = "en"

## sources

- https://github.com/project/blob/main/README.md
- docs/architecture.md
- README.md

## audio

enabled = true
description = "In-depth discussion of technical architecture and design patterns"
format = "DEEP_DIVE"
length = "DEFAULT"

## video

enabled = true
description = "Introduction tutorial for beginner developers"
format = "EXPLAINER"
style = "WHITEBOARD"

## report

enabled = true
format = "STUDY_GUIDE"
append = "Include complete code examples and best practices"

## slide_deck

enabled = true
description = "Technical sharing presentation"
format = "PRESENTER_SLIDES"
length = "DEFAULT"

## infographic

enabled = true
description = "Project architecture component relationship diagram"
orientation = "LANDSCAPE"
detail = "STANDARD"
style = "PROFESSIONAL"

## quiz

enabled = false
difficulty = "MEDIUM"
quantity = "STANDARD"

## flashcards

enabled = true
description = "Key concepts and API study cards"
difficulty = "MEDIUM"
quantity = "STANDARD"

## mind_map

enabled = true

Output File Mapping#

Generated artifacts will be downloaded to the outputs/ directory with naming conventions as follows:

Important Notes#

1. Language is global - Language settings affect all artifacts and cannot be set independently per artifact

2. enabled flag - Only artifacts with enabled = true will be generated; default is false

3. Multiple artifact types - You can declare multiple artifact blocks of the same type with different parameters. Each enabled block will generate independently. For example, you can generate multiple videos with different styles.

4. Source priority - By default, the software will upload all files in the material directory (except the outputs directory and notebooklm.md file). If you want to add web links, you can include them here.

5. Advanced features - Some features like cinematic_video require Google AI Ultra subscription

Quick Reference#

Common Combinations#

Learning material generation:

• Audio + Report + Flashcards + Quiz
• Suitable for quick understanding and learning

Content marketing:

• Video + Infographic + Blog Post
• Suitable for content creators and social media

Technical documentation:

• Report (STUDY_GUIDE) + Slide Deck + Mind Map
• Suitable for team training and knowledge preservation

Presentation preparation:

• Audio + Slide Deck + Infographic

• Suitable for conference presentations and public speaking

Troubleshooting#