Skip to content
DUETA
Sign upConsoleLogin

Separate the Inseparable.

One API for all audio separation.

Get started

Separate your voices

Mixed
separated
Speaker 1
Speaker 2
Open in console →
  • Original0:07
    Speaker 10:07
    Speaker 20:07
    DuplexDuplex Conversation SeparationTwo people talking over each other, returned as two clean tracks.Run the sample
    • Two speakers, two tracks
    • Word timing kept
    • Room tone preserved
  • As recordedSeparated
    Mic 1
    Mic 2
    Mic 3
    Mic 4
    Mic 5
    Mic 6
    MicsDominant Audio SeparationRecordings from one room, each returned holding its own source.Clean the mics
    • Every mic hears everyone
    • Its own source kept, bleed gone
    • The room kept as its own bed
  • LiveRealtime Speaker ExtractionComing soonOne voice lifted out of a live stream as it arrives.
    • Separates as it arrives
    • One voice isolated
    • The same API as Duplex
  • StemsMusic Source SeparationComing soonA mix pulled apart into vocals, drums, bass and the rest.
    • Vocals and instrumental
    • Vocals, drums, bass and the rest — as separate stems
    • Lossless output
DUETA engines
    1. The take, as recorded
    2. Noise removal
    3. Enhancement
    4. Speaker separation
    5. Verification
    1. Enroll the voice
    2. A live stream arrives
    3. Only the enrolled voice remains
    1. Where each part sits
    2. Mid-Side Analysis
    3. Artist-ready stems
    1. One room, a mic each
    2. Every mic hears everyone
    3. Only the dominant signal remains

Now, with Your Audio.

Every engine, one API, same three calls.

Get startedAPI docs