- Home
- Stem Separation Software
- StemDeck
StemDeck
What it is
StemDeck separates songs into individual audio stems for practice, study, transcription, remixing, and other creative work. Its Demucs htdemucs_6s model can split vocals, drums, bass, guitar, piano, and other sounds. It accepts MP3, WAV, FLAC, OGG/Opus, MP4, and M4A files, along with YouTube URLs. In the workspace, waveform lanes have mute and solo controls, volume adjustment, and looping; users can export individual stems or a custom mix in WAV or MP3. Song analysis includes BPM, key, scale, confidence, integrated LUFS, and sample peak level. StemDeck says processing happens locally and that the app does not transmit, store, cache, or redistribute audio. Internet access is needed for YouTube downloads and an initial model fetch of about 170 MB; the model is cached afterward. The project documents Docker and Unraid deployment, running its web server from source, and a local REST API. StemDeck is free, with no account, quota, uploads, or subscription, but it processes one job at a time and supports up to six stems. Processing speed depends on local hardware, and there is no mobile app.
For and against
For
- Free, with no account or subscription
- Separates up to six stem types
- Supports common audio formats and YouTube URLs
- Exports stems or mixes in WAV and MP3
- Processing runs locally, according to the project
Against
- Processes one job at a time
- No mobile app
- Processing speed depends on local hardware
- Initial model fetch needs about 170 MB of internet
The verdict
StemDeck combines stem separation with editing, analysis, and export tools in a free project. Its one-job limit and dependence on local hardware are important constraints.
Who it suits
StemDeck is aimed at musicians, producers, teachers, hobbyists, and listeners who want to practice, study, or remix songs. It may suit people who can run processing on local hardware and do not need a mobile app.
Plans and prices
StemDeck
Free, forever · no account, quota, uploads, or subscription · one job at a time
Read 30 Sep 2026 · github.com
On the record
- API
- The local server exposes REST endpoints for jobs, status, cancellation, stem streaming, and video export, plus server-sent job events.github.com · read 30 Sep 2026
- Audience
- StemDeck is aimed at musicians, producers, teachers, hobbyists, and listeners who want a focused tool for practice, study, or remixing.stemdeck.app · read 30 Sep 2026
- Audio analysis
- Song analysis includes BPM, key, scale, confidence, integrated LUFS, and sample peak level.github.com · read 30 Sep 2026
- Import formats
- It accepts MP3, WAV, FLAC, OGG/Opus, MP4, and M4A files, as well as YouTube URLs.github.com · read 30 Sep 2026
- Internet needs
- Internet is needed for YouTube downloads and the initial model fetch of about 170 MB; the model is cached afterward.github.com · read 30 Sep 2026
- Mixer and editing
- The workspace includes waveform lanes, mute and solo controls, volume adjustment, looping, and exports of individual stems or a custom mix.github.com · read 30 Sep 2026
- Notable limits
- The project describes one job at a time, up to six stems, and no mobile app; it also says processing speed depends on local hardware.github.com · read 30 Sep 2026
- Privacy
- StemDeck says audio processing runs locally and audio is not transmitted, stored, cached, or redistributed by the app.github.com · read 30 Sep 2026
- Purpose
- StemDeck separates songs into stems for practice, study, transcription, remixing, and creative audio workflows.github.com · read 30 Sep 2026
- Self-hosting
- The project documents Docker and Unraid deployment as well as running the web server from source.github.com · read 30 Sep 2026
- Stem types
- The Demucs htdemucs_6s model separates vocals, drums, bass, guitar, piano, and other sounds.github.com · read 30 Sep 2026
- Support and project
- The maker links to GitHub issues and pull requests, Discord, Reddit, Instagram, and X; the project is licensed under Apache-2.0.github.com · read 30 Sep 2026
