DonutMediaCenter

A local library for images, videos, and music, with automatic image tagging and browsing built around each media type.

DonutMediaCenter keeps the existing LocalBooru application identifier and data directory so current libraries, favorites, and settings remain available after upgrading. Published releases may still carry the LocalBooru filename until the next release is built.

- Help, feature requests, and discussions

🌐 Visit Website

Download

Platform	Download
Windows (legacy 0.3.33)	Portable (ZIP) | Portable (EXE) | Installer
Linux	Portable (ZIP) | AppImage | .deb | .rpm
macOS (legacy 0.3.33)	Portable (ZIP) | DMG (Apple Silicon)

The macOS links are legacy compatibility builds, not a current stable support claim. New native artifacts remain inspection-only until the exact package passes database, gallery, video/sidecar, import, and sustained-use acceptance on real Apple hardware. See docs/agents/release-and-infrastructure.md.

USB Portable Mode

The Portable versions automatically store all data next to the app - no setup needed:

1. Download and extract to your USB drive or any folder
2. Run the app - a data folder is created automatically
3. Copy the entire folder to any computer and your library comes with you

The core library and local music metadata indexing do not require Python. Optional add-ons may have their own dependencies.

The installer versions use AppData/home directory instead (shared across updates).

Features

Media Libraries

• Images, Videos, and Music - Browse each media type with its own filters and collections
• Albums and Songs - Play album tracks in disc order or start a song-based related mix
• Local related listening - Queue similar songs from your own library without a streaming service
• Persistent player - Keep music playing while browsing images, videos, and collections
• Existing organization - Legacy image and video collections keep their members after the galleries are separated

Automatic AI Tagging

• Booru-style tags - Automatically tags images with character, artist, copyright, and general tags using AI models
• Face detection - Detects faces in images and estimates age
• Content rating - Automatically classifies images by rating (PG, PG13, R, X, XXX)

Smart Organization

• Watch directories - Add folders containing images, videos, or audio; the library indexes new files as they appear
• Tag-based search - Find images instantly using tag combinations
• Favorites - Mark your best images and filter to show only favorites
• Pruning - Clean up non-favorited images by moving them to a dumpster folder

Modern Interface

• Masonry grid - Beautiful responsive gallery layout
• Media viewers - Image navigation, video playback controls, and an album or song queue for music
• Dark theme - Easy on the eyes for extended browsing sessions
• Relevant filters - Filter images, videos, and music using the information available for each type

Flexible Access

• Desktop app - Native Tauri application for Windows, macOS, and Linux
• Web browser - Access your local server from a browser when network access is enabled
• System tray - Runs quietly in the background, always ready

Auto Updates

• Automatically checks for new versions
• One-click update installation

Installation

Requirements

• Windows/macOS: Use a supported desktop build.
• Linux: Use a supported package or AppImage; optional add-ons may require extra dependencies.

Getting Started

1. Launch DonutMediaCenter - The app starts with an empty library
2. Add a directory - Go to "Directories" and add folders containing images, videos, or music
3. Wait for indexing - DonutMediaCenter scans media and runs enabled image tagging add-ons
4. Browse and listen - Use Images, Videos, or Music to search, organize, and play your library

Usage Tips

Searching

• Click tags in the sidebar to filter images
• Use the search box to find specific tags
• Combine multiple tags to narrow results

Keyboard Shortcuts (Lightbox)

• ← → - Navigate between images
• Escape - Close lightbox
• F - Toggle favorite

Filtering

• Ratings - Toggle rating buttons to show/hide content levels
• Age range - Filter by detected face age
• Favorites only - Show only your favorited images
• Directory - Filter to a specific watched folder

Pruning

Keep your collection clean by pruning directories:

1. Go through images and favorite the ones you want to keep
2. Go to "Directories" and click "Prune" on a folder
3. Non-favorited images are moved to the dumpster folder

Development

Prerequisites

• Node.js 20+
• Python 3.10+
• pip

Setup

# Clone the repository
git clone https://github.com/DonutsDelivery/LocalBooru.git
cd LocalBooru

# Install Node dependencies
npm install

# Install Python dependencies
pip install -r requirements.txt

# Build the frontend
npm run build:frontend

# Run in development mode
npm run dev

Building

# Build for current platform
npm run build

# Build for specific platforms
npm run build:linux
npm run build:win

Tech Stack

• Frontend: React + Vite
• Desktop: Tauri (Rust) / Electron (legacy)
• Backend: Python FastAPI
• Database: SQLite
• AI/ML: ONNX Runtime, InsightFace
• Video: GStreamer

Building from Source

Tauri (Recommended)

# Build all verified Linux release artifacts with Docker/Podman
npm run release:linux

# Or build specific formats
npm run tauri:build:linux:appimage
npm run tauri:build:linux:deb
npm run tauri:build:linux:rpm

See docs/LINUX_PACKAGING.md and the release infrastructure guide for artifact names, persistent cache configuration, and verification gates. The retired Electron workflow must not be used for current releases.

License

MIT License - See LICENSE for details.

Credits

Built with love for image collectors everywhere.