OpenHomeKaraoke
github.com
- Free plan
- Windows
- macOS
- Linux
- Web
- Self-hosted
What OpenHomeKaraoke is
OpenHomeKaraoke is a free, open-source home karaoke system that searches for songs online, downloads them, and manages a shared queue. It can search YouTube, Youku, Bilibili, and other video sites, while playing MP3 and CDG files, including ZIP bundles. A DNN vocal splitter offers instrumental, vocal, or combined playback and requires PyTorch, preferably with GPU support. Voice search and queueing use OpenAI’s multilingual Whisper model. People on the same network can use a phone or computer browser to search the local library and add songs. Playback includes pause, skip, restart, seeking, volume, pitch, tempo, and audio or subtitle delay controls. Karaoke can appear on a TV through HDMI, screen sharing, or an HTTP stream viewed in a browser. It supports Linux, Windows, macOS, Raspberry Pi, web, and self-hosted use. Setup requires Python dependencies, VLC, and ffmpeg.
Who it suits
It suits households or local groups that want browser-based shared karaoke with online song search and multiple playback controls. Users with suitable hardware can add vocal splitting, voice search, and TV display options.
Good to know, and what to watch
- Free and open source
- Searches multiple video sites
- Shared queue for network users
- Voice search with Whisper
- Supports MP3, CDG, and ZIP files
- Offline mode cannot search or download
- Vocal splitting requires PyTorch
- Large libraries may strain some devices
- No GitHub releases are listed
The verdict
OpenHomeKaraoke provides a broad self-hosted karaoke setup with shared browsing, online downloads, vocal options, and TV output. Hardware requirements and offline limitations matter for larger libraries or advanced processing.
What the maker publishes
- Access control
- An optional administrator password can restrict queue editing, player controls, song editing, and system shutdown; without one, everyone is an admin. source
- Admin controls
- An optional administrator password can restrict queue editing, player controls, song editing and system shutdown; without it, everyone is an admin. source
- Device limits
- The README says concurrent downloads and browsing large song libraries may suffer on some devices, and recommends multi-core Raspberry Pi models. source
- Display
- The player can display on a TV via HDMI or screen sharing, and can stream its captured screen over HTTP for viewing in a browser. source
- License
- The repository identifies its license as GPL-3.0. source
- License and releases
- The repository identifies a GPL-3.0 license and its GitHub releases page says there are no releases. source
- Media formats
- The project supports MP3 and CDG files, including compressed ZIP bundles. source
- Offline use
- Without a network connection, users can access the existing song library but cannot search or download new songs from the internet. source
- Performance limit
- The README cautions that concurrent downloads and browsing large song libraries may suffer, and recommends multi-core Raspberry Pi devices. source
- Playback
- It supports mp3 and cdg files, including compressed .zip bundles. source
- Playback controls
- The interface supports pause, skip, restart, volume, pitch and tempo changes, seeking, and audio and subtitle delay adjustments. source
- Purpose
- OpenHomeKaraoke is an open-source, YouTube-based home karaoke system for searching, downloading and queueing songs. source
- Requirements
- Installation instructions call for Python dependencies, VLC and ffmpeg, and describe GPU setup for accelerated vocal splitting or speech recognition. source
- Security status
- GitHub reports that the repository has no detected security policy and no published security advisories. source
- Shared queue
- Multiple users on the same network can use a smartphone or computer browser to search the local library and queue songs. source
- Shared web interface
- Multiple users on the same network can use a browser, including a smartphone browser, to search the local library and queue songs. source
- Song discovery
- Users can search YouTube and download songs from YouTube and other video sites, including Youku and Bilibili. source
- Song sources
- Users can search YouTube and download songs from YouTube and other video sites, including Youku and Bilibili. source
- Speech recognition
- Users can search for and queue songs by voice using OpenAI's multilingual Whisper speech recognition model. source
- Supported systems
- The README says it works on Linux, Windows, Raspberry Pi, and Mac OS. source
- TV display
- The karaoke display can be shown on a TV over HDMI, screen sharing, or a TV web browser using an HTTP stream. source
- Vocal options
- A DNN-based vocal splitter can provide instrumental, vocal, or combined playback; the project also describes stereo track subtraction as an alternative method. source
- Vocal separation
- A DNN-based vocal splitter can play instrumental, vocal or both tracks; it requires PyTorch, preferably with GPU support. source
- Voice search
- Voice-based song search uses OpenAI’s multilingual Whisper speech recognition model. source
Checked against the maker's pages on 8 Oct 2026.