A minimalistic note taking application for creative humans that use computers made by some weirdo w/Claude Opus 4.5
- Free- no account or subscription required
- Works offline (only needs to be online once to make image canvas work)
- Doesn't manage syncing notes (use Dropbox/SyncThing/OneDrive/Google Drive/whatever to manage it yourself
- Image attachments - Attach images to any note (PNG, JPG, GIF, WebP) via drag/drop from a browser or whatever
- Audio attachments - Attach and play audio files with built-in player (supports MP3, WAV, and more with automatic transcoding)
- Built-in drawing pad - Create hand-drawn sketches and diagrams with Fabric.js canvas
- PDF Document - Export all notes in a folder as a single PDF
- PNG Images - Export all notes as individual PNG images
- EPUB3 Book - Create an e-book from your notes
- HTML Website - Export individual notes as standalone web pages
-
GitHub Pages publishing - Publish notes directly to your GitHub repository
-
** For your "ahoy" page on your published site just publish a note titled "ahoy" (without the "")
-
Cryptocurrency tipping - Add Bitcoin and Ethereum tip addresses to published notes
-
Custom publishing name - Brand your published notes with your name
- Smart text correction - Fix spelling, grammar, and typos using AI
- Two correction modes:
- Manual mode - Select text and click fix to correct only the highlighted portion
- Auto mode - Automatically corrects newly typed text (tracks dirty regions for efficiency)
- Multiple providers:
- OpenAI - Uses GPT-4o-mini for fast, accurate corrections
- Local LLM - Use your own GGUF model files for offline correction
- ** A small, fast well balanced Local LLM to use is gemma-2-2b-it-Q4_K_M.gguf
- Preserves formatting - Respects line breaks, paragraphs, and document structure
- drag/drop notes from the note list into the active note- then double click the [[note name]] to go to that note
- drag/drop files from the Operating system into the active note- then double click the path text to go to that file
| Shortcut | Action |
|---|---|
Arrow Up/Down |
Navigate note list |
Spacebar |
Toggle Select/Draw mode (on canvas) |
Delete/Backspace |
Delete selected canvas objects |
Escape |
Close modals/lightbox |
- Node.js 18+ installed
- npm or yarn
cd noatboat-electron
npm installnpm startWindows:
npm run build-winMac:
npm run build-macLinux:
npm run build-linuxAll platforms:
npm run build-allBuilt executables will be in the dist/ folder.
- Click the folder icon (📁) to select a folder for your notes
- The folder path is saved automatically - next time you launch, it loads automatically
- Create new notes with the new note button (📄)
- Notes are saved as
.txtfiles in your selected folder - Navigate subfolders with the folder navigation buttons
- Attach images (🖼️), audio (🔊), or create drawings (✏️)
- Export or publish your notes using the export button (📤)
your-notes-folder/
├── note-title.txt # Note content
├── note-title.png # Image attachment (or .jpg, .jpeg, .gif, .webp)
├── note-title.canvas.json # Canvas/drawing data
├── note-title.canvas.png # Canvas preview image
├── note-title.audio.mp3 # Audio attachment
└── subfolder/ # Nested folders supported
└── another-note.txt
Access preferences via Cmd/Ctrl + , or the app menu:
- Theme - Light or Dark mode
- Focus Mode Strength - Adjust how much the UI dims in focus mode
- Text Correction - Off, Auto, or Manual mode
- Auto-Fix Provider - OpenAI or Local LLM
- OpenAI API Key - For cloud-based text correction
- Local Model Path - Path to your GGUF model file
- GitHub Token - For publishing to GitHub Pages
- GitHub Repository - Target repo for publishing (format:
username/repo) - Publishing Name - Your name for published notes
- Tip Addresses - Bitcoin and Ethereum addresses for tips
The app stores its configuration in:
- Windows:
%APPDATA%/noatboat/config.json - Mac:
~/Library/Application Support/noatboat/config.json - Linux:
~/.config/noatboat/config.json
- Electron - Cross-platform desktop framework
- Fabric.js - Canvas drawing and manipulation
- FFmpeg - Audio transcoding (bundled)
- node-llama-cpp - Local LLM inference (optional)
- OpenAI API - Cloud text correction (optional)
GPLv3