What it does
Every track, taken apart.
Stem separation runs on your GPU the moment you add a track — then it's yours to isolate, mute, mash, and mix.
AI stem separation
An AI model splits each track into drums, bass, melody and vocals. Solo or mute any stem, or lay one deck's acapella over the other's instrumental.
Library that sorts itself
Every track is analyzed for BPM, musical key (Camelot) and time signature. Load one deck and the library re-ranks by what mixes best — same key, adjacent key, closest tempo.
Beatmatch & lock
Octave-aware sync (80 ↔︎ 160 just works), snap-to-beat on the beat grid, and a sync-jog that shuttles both decks forward or back while staying phase-locked.
A real deck
Hot cues, beat loops, rolls and beat-jump; 3-band EQ + filter; fine tempo trim, type-in BPM and key-shift. Pre-listen in your headphones on a separate output — and map every control to a MIDI controller or the keyboard.
Capture the set
Record the master mix — crossfades, EQ, mashups and all — straight to a file you can download.
Three looks, dark & light
Minimal-techno white-label, a VHS mixtape, or a photocopied rave flyer — each in light and dark. Purely cosmetic, instantly switchable.
Private by design
It all runs on your machine.
There's no server. Tracks are separated once with WebGPU, then cached in your browser for instant reuse next session. Your music never leaves your computer.
- 01Add tracks — they're separated & analyzed on your GPU, then cached.
- 02Load two onto the decks; the library shows your best next mix.
- 03Sync, beatmatch, ride the EQ and stems — then hit record.