Isolate vocals from a track to create an acapella or instant instrumental
Mix two tracks together with automatic BPM detection and beat sync
Change a track's tempo without changing its pitch using keylock mode
Organize a local music library into crates and record a finished mix
| alexanderliteplo/stemdeck | 4ssh1/mine-sweeper | achawla19/intuitcode-extension | |
|---|---|---|---|
| Stars | 2 | 2 | 2 |
| Language | TypeScript | TypeScript | TypeScript |
| Last pushed | — | 2026-01-31 | — |
| Maintenance | — | Maintained | — |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 3/5 | 2/5 | 4/5 |
| Audience | vibe coder | vibe coder | developer |
Figures from each repo's GitHub metadata at analysis time.
Stem splitting needs a separate audio-separator install and a Python version of 3.13 or lower.
StemDeck is a desktop DJ application that runs entirely on your own computer, using only the music files you already have. It gives you two decks and a mixer, similar to what a physical DJ setup offers, plus a feature that most consumer DJ software does not have: it can split a track into separate parts, such as vocals, drums, bass, and everything else, using AI models that run locally. That means you can mute the vocals to get an instant instrumental, isolate just the drums, or play an acapella from one song over the beat of another, all without sending your music anywhere online. On top of the stem splitting, StemDeck includes the tools a DJ expects: a three band equalizer per deck, a filter knob, a crossfader, automatic beat and tempo detection with a sync button to match two tracks, pitch and tempo adjustment that can keep the pitch steady while the tempo changes, hot cues, loops, waveforms you can see and click on, and the ability to record your mix to a file. Tracks can be organized into folders called crates inside a simple two pane library view. The stem splitting itself is handled by a separate open source tool called audio-separator, which downloads its AI models the first time you use them and then works offline from then on. Split versions of a track are saved so a track only needs to be processed once. Users on Apple Silicon or with an NVIDIA graphics card can install a faster version of this tool. Everything else in the app works even without it installed, you just cannot split new tracks until it is set up. The app itself is built with Electron, React, and TypeScript, and processes audio using the browser's built in Web Audio tools. It is offered under the MIT license, with one supporting library used for the pitch preserving tempo feature licensed separately under LGPL.
A local desktop DJ app with two decks, a full mixer, and AI powered stem splitting that isolates vocals, drums, and bass from any track on your computer.
Mainly TypeScript. The stack also includes TypeScript, Electron, React.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice, one bundled audio library is separately licensed under LGPL.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.