Demixr Pro audio separation suite
Separate a song into audio you can actually use.
Demixr gives creators, mixers, and content teams vocal separation, dereverb, VAD markers, BPM/key analysis, and stem separation. It works like a focused audio workstation, not a disposable utility.

26.93 dB
Internal average vocal SDR
6
Audio workflows in one place
500MB
Pro per-upload limit
Core tools
Separate, clean, segment, and analyze in one workflow.
Real listening tests
Hear the output before uploading your own audio.
The lab page includes original, vocal, and instrumental outputs from local samples. The homepage highlights a few vocal results so users can judge model character quickly.
Harmony separation
VocalComplex separation experiment
VocalNoise separation experiment
VocalProduction-ready export flow
VAD markers, batch export, and sampling settings are designed for real workflows.
Clear retention policy
Uploaded files follow backend cleanup policy for teams that need clear data boundaries.
Continuously improved Pro model
The lab keeps publishing real sample results across difficult audio conditions.
Start with your first track.
Demix, dereverb, VAD, BPM, and lab results are ready.