What does a voice isolator do?

It separates the voice in a recording from everything playing behind it, and returns both parts. It does not repair a recording: clipping, distortion and heavy room noise stay in the isolated voice.

Does it work on speech, not just singing?

It works on clear speech over music, which is the common case for podcasts, interviews and video. The model was trained on sung voice in music, so speech buried under a dense mix separates less cleanly than singing does.

Is my recording uploaded?

No. Isolation runs inside this browser tab: the open-source model is downloaded once and cached, then your recording is processed on your own machine. Nothing is uploaded, nothing is stored, and your audio is never used to train ANYANO models.

Can it remove background noise?

Only the musical background. It separates voice from music, not voice from traffic, hum or fan noise — those are a different kind of model, and this page will not pretend otherwise.

100 MB and 12 minutes per file, six separations an hour per visitor, in MP3, WAV, FLAC, M4A, AAC or OGG.

Free tool

AI Voice Isolator

Separate a voice from the music behind it, and hear both halves on their own before you download either.

Isolation runs inside this browser tab, not on our servers. Your recording never leaves your machine and is never used to train ANYANO models.

Choose an audio file to separate

MP3, WAV, FLAC, M4A, AAC or OGG — up to 100.0 MB and 8 minutes. Processed on your own machine.

The model was trained on sung voice in music, so it handles singing and clear speech over music well, and does not repair damaged or clipped recordings.

Detail

How this works

Isolation works by prediction, not filtering. The model estimates the voice and the backing as two separate signals, so where they overlap in frequency and time the split becomes a judgement call. A voice sitting clearly above sparse music isolates almost transparently; a voice competing with a full mix leaves artifacts.

Both results come back at the original length and position, which makes the comparison player the right way to judge them: play everything together, then solo the voice.

This tool cleans up existing recordings. ANYANO does the opposite — it trains on voices and songs you own so you can generate new material in your own direction.

FAQ

Frequently asked questions

Toolbox

See all free ANYANO tools

Train your sound. Generate your song.

ANYANO is the bigger version of this: prepare music you own, train a Sound that stays private to your account, then generate from it.