Upload a track and separate it right here in your browser — no upload to any server, nothing leaves your device.
Higher = cleaner separation but a little more processing artifact. Move the slider, then it re-separates.
Real AI model, runs 100% on your device — no server. First time downloads ~170 MB (use WiFi), then processes in a few minutes. Best on a computer or newer phone.
Cleanest on stereo studio recordings. Output is full-resolution 24-bit WAV.
Everything runs locally — your audio never leaves your device.