If you have used Google NotebookLM, you might have noticed some limitations when processing e-books. This tool is designed to address those pain points, offering a more professional and flexible e-book reading and mind-mapping experience:

Feature	This Tool (ebook-to-mindmap)	Google NotebookLM
Mind Map Export & Editing	🛠️ Export for Editing: Although the online interface is read-only, you can export the mind map as a data file and import it into the Mind Elixir Desktop Client or other editors for full editing and modification.	❌ No Export or Editing: Mind maps are only for read-only viewing, cannot be exported, and do not support any post-editing.
Prompt Flexibility	🎯 Total Customization: Supports custom prompts and offers a "Use Custom Prompt Only" mode to completely override system prompts. AI fully respects your instructions.	🚫 Overpowering System Prompts: Primarily optimized for retrieval and summarization (RAG). Its built-in system prompts are highly dominant, making it unable to fully follow your own custom instructions.

✨ Features

📚 Multi-format Support

EPUB Files: Complete support for parsing and processing EPUB format e-books
PDF Files: Intelligent PDF document parsing with table of contents-based and smart chapter extraction

🤖 AI-Powered Content Processing

Multiple AI Services: Support for Google Gemini and OpenAI GPT models
Dual Configuration Modes: Support for configuring your own API Keys (BYOK Mode), or logging into your Mind Elixir account to use built-in models (No Key Option)
BYOK Mode: Use your own API Key (Bring Your Own Key), where all AI requests are made directly from your browser to the AI providers, ensuring complete data security and privacy
No Key Option: Directly log in with a Mind Elixir account to use built-in hosted models (consuming Star Power/星星算力). Out-of-the-box, no API Key configuration required
Three Processing Modes:
- 📝 Text Summary Mode: Generate chapter summaries, analyze chapter relationships, output complete book summary
- 🧠 Chapter Mind Map Mode: Generate independent mind maps for each chapter
- 🌐 Whole Book Mind Map Mode: Integrate entire book content into one comprehensive mind map

💾 Efficient Caching Mechanism

Smart Caching: Automatically cache AI processing results, allowing continuation from the last position if processing is interrupted
Cache Management: Support clearing cache by mode to save storage space
Offline Viewing: Processed content can be viewed offline

🎨 Modern Interface

Responsive Design: Adapts to various screen sizes
Real-time Progress: Visualized processing with real-time step display
Interactive Mind Maps: Support zooming, dragging, node expand/collapse
Content Preview: Support viewing original chapter content

📖 User Guide

1. Configure AI Service

First-time use requires AI service configuration, supporting the following two options:

Option 1: BYOK Mode (Bring Your Own Key)

🔒 Privacy Protection Note: This tool supports using your own API Key (Bring Your Own Key). All AI requests are made directly from your browser to the AI providers (Google or OpenAI), never through any third-party proxy or intermediary servers, ensuring your data security and privacy.

Click the "Configure" button
Select AI service provider, Google Gemini is recommended for trial
Enter the corresponding API Key
Fill in the model

Getting API Keys

Using Google Gemini as an example:

Visit Google AI Studio
Sign in with your Google account
Create a new API Key
Copy the API Key to configuration

For more information on AI provider options, refer to: Free and Paid AI API Selection Guide

Option 2: No Key Option (Star Power / 星星算力) 🌟 Recommended

If you don't have your own AI provider's API Key or prefer to skip complex configurations, you can log in with your Mind Elixir account to use the built-in hosted models (Star Power):

⚡ Out-of-the-box: No API Key configuration required. Simply log in and use.
🪙 Pay-as-you-go: Deducted from your Star Power balance.
🔄 Cross-app Sharing: Star Power balance is shared across the entire Mind Elixir app ecosystem (e.g., it is also shared in the M10C Browser Extension).
⏳ Never Expires: Star Power balances purchased or obtained remain valid permanently with no time limit.

To recharge or manage your Star Power, please visit the Mind Elixir Website.

2. Upload E-book File

Click "Select EPUB or PDF File" button
Choose the e-book file to process
Supported formats: .epub, .pdf

You can get free e-books from websites like Project Gutenberg and Standard Ebooks.

3. Configure Processing Options

Set processing parameters in the configuration dialog:

Processing Mode

Text Summary Mode: Suitable for scenarios requiring text summaries
Chapter Mind Map Mode: Generate independent mind maps for each chapter
Whole Book Mind Map Mode: Generate unified mind map for the entire book (may fail due to insufficient model context if the book content is too long)

Book Type

Fiction: Suitable for novels and story books
Non-fiction: Suitable for textbooks, reference books, technical books, etc.

Advanced Options

Smart Chapter Detection: When enabled, uses AI to intelligently identify chapter boundaries
Skip Irrelevant Chapters: Automatically skip prefaces, table of contents, acknowledgments, etc.
Sub-chapter Depth: Set the hierarchy depth for extracting sub-chapters (0-3)

4. Extract Chapters

Click "Extract Chapters" button
System will automatically parse the file and extract chapter structure
After extraction, chapter list will be displayed
You can select chapters to process

Chapter Grouping (Optional)

For books with comprehensive content, you can merge multiple small chapters into a group for processing, allowing for a more structured result. By grouping, AI will analyze the content of these chapters together.

Select Chapters: Click on a chapter list item (not the checkbox area) to select it. Multiple selection is supported. Selected chapters will have a highlighted border.
Add Tag:
- Click the "Add Tag" button at the top of the list
- Or use the shortcut Ctrl + G
Enter Group Name: In the dialog that appears, enter a group name (e.g., "Part 1", "Chapter 1", etc.). After confirmation, the selected chapters will be marked as the same group.
Remove Group: Click the "X" icon after the existing tag on the chapter list to remove the grouping for that chapter.

4.5. Custom Prompts (Advanced)

You can manage your dedicated prompts on the "Custom Prompts" page and use them when processing books to get results that better fit your needs.

Manage Prompts:
- Click "Custom Prompts" in the navigation bar to enter the management page
- Click "Add Prompt" to create a new prompt template
- Fill in the name, description, and specific prompt content
- Support editing, copying, and deleting existing prompts
Use Prompts:
- At the bottom of the configuration interface after extracting chapters, find the "Custom Prompts" option
- Select your pre-created prompt from the dropdown menu
- Use Custom Prompt Only: When checked, your prompt will completely replace the system default prompt (this option only applies to Text Summary Mode). If unchecked, your prompt will be sent to the AI as supplementary instructions.

5. Start Processing

Confirm selected chapters
Click "Start Processing" button
System will display processing progress and current steps
Results will be shown after completion

6. View Results

Depending on the selected processing mode, you can view different types of results:

Text Summary Mode

Chapter Summaries: Detailed summary of each chapter
Chapter Relationships: Analysis of logical relationships between chapters
Book Summary: Core content summary of the entire book

Mind Map Mode

Interactive Mind Maps: Zoomable, draggable mind maps
Node Details: Click nodes to view detailed content
Export Function: Support exporting as images or other formats

🛠️ Technical Architecture

Core Technology Stack

Frontend Framework: React 19 + TypeScript
Build Tool: Vite
Styling: Tailwind CSS + shadcn/ui
State Management: Zustand
File Parsing:
- EPUB: @smoores/epub + epubjs
- PDF: pdfjs-dist
Mind Maps: mind-elixir
AI Services:
- Google Gemini: @google/generative-ai
- OpenAI: Custom implementation

📄 License

This project is licensed under the MIT License. See LICENSE file for details.

🙏 Acknowledgments

Thanks to the following open source projects:

For questions or suggestions, please submit an Issue or contact the developer (WeChat👇)