Vocal Isolator: isolate the vocals from a song.
- Vocal + music as two files
- Price shown before upload
- Ready in a few minutes
- Failed splits refunded
Use only with music you own or have permission to use.
Example result
In short
A vocal isolator separates the voice from the music in a recording. Drop your file above and an AI model on a cloud GPU returns the vocal on its own and the music without it as two files, usually in a few minutes. It costs 4 credits per song (about $0.60 or less), and failed splits are refunded automatically.
How it works
How to use the vocal isolator
- 1
Drop the recording
Isolate vocals from song demos, finished mixes or narration you own or have permission to use. MP3 or WAV, up to 10 minutes — a WAV over 20 MB is compressed to a 256–320 kbps MP3 in your browser first.
- 2
Isolate the vocals
Press Split (4 credits). The price is on the button before you confirm, and the vocal isolator separates the voice from the music on a cloud GPU in a few minutes.
- 3
Download the vocal
Solo the voice in your browser and download both files. Links stay valid for about a week, so save what you need.
What a vocal isolator does — and what it doesn't
A vocal isolator separates the voice from the music. You get two files of the same length: the vocal on its own — lead, harmonies and ad-libs — and everything else without it.
What it doesn't do is clean up noise. Hum, hiss, wind and room echo aren't music, so the model leaves most of them with the voice. For that, you want a dialogue noise reducer rather than a vocal isolator.
- Separates: lead and backing vocals, rap, spoken word over music
- Keeps with the voice: the EQ, compression and part of the reverb from the mix
Who uses a vocal isolator
Same split, different reasons — always on recordings you have the rights to.
Singers learning a part
Hear every breath, slide and harmony without the band in the way. Loop a phrase in your DAW until it sits in your voice.
Cover artists
Study the original phrasing, then record your own take over the music from the same split. For a karaoke backing track, see the instrumental maker.
Podcasters and video creators
Lost the separate tracks of an old episode with music under the voice? The vocal isolator works as a voice extractor: you get a speech track to re-edit and re-score, up to 10 minutes per file.
Remixers and producers
Start a new production from your own song's vocal. For DJ edits and sampling, the acapella extractor goes deeper.
How the AI vocal isolator works
AI vocal isolation doesn't rely on stereo tricks like phase cancellation: the model was trained on multitrack recordings where it could hear voices and instruments apart, so it recognizes what a voice sounds like — vowels, consonants, breaths, vibrato. Moment by moment, the vocal isolator separates the voice into one file; everything else becomes the music file.
Some apps call this a vocal extractor; it is the same split, and it runs on a cloud GPU rather than in your browser.
Vocal isolator results by material
The recording matters more than any setting. This is what to expect.
| Material | Typical result | What to watch for |
|---|---|---|
| Modern pop, hip-hop, EDM | Very clean | Faint hi-hat ghosts on the "s" sounds |
| Acoustic singer-songwriter | Clean | Hard-strummed guitar can bleed into loud phrases |
| Dense rock or metal | Usable | More bleed from distorted guitars and cymbals |
| Live recordings | Usable, varies | Crowd and room sound stay around the voice |
| Speech over a music bed | Usually clean | Background noise and room echo stay with the voice |
Typical results, not guarantees.
Vocal isolator, acapella extractor, vocal remover: which page?
The first four pages run the same 2-stem split and give you both files; they differ in focus.
| Page | Focus | Credits |
|---|---|---|
| Vocal Isolator | The voice on its own — learning, covers, podcasts | 4 credits |
| Acapella Extractor | Acapellas for DJ edits, remixes and sampling | 4 credits |
| Vocal Remover | The song without vocals | 4 credits |
| Instrumental Maker | Karaoke and rehearsal tracks | 4 credits |
| Stem Splitter | Vocals, drums, bass, other + up to 12 parts, depending on the song | 7 credits |
Use the vocal isolator for covers and practice
The quickest way to learn a vocal is to hear it alone. Load the isolated vocal and the music into your DAW at bar 1, loop the hard phrase and sing along until the phrasing feels natural. Then mute the guide and sing over the music alone.
Check your range first with the key finder. Working on a Suno or Udio song? Isolate the vocal to re-record or rework it, then finish the track with Suno mastering.
- Loop phrases at full speed first, then slow them down in your DAW
- Keep the original vocal 10–15 dB under your own as a guide
Tips to isolate vocals cleanly
A few habits before and after the split make the vocal easier to use.
- Start from your best source. A WAV master lets you extract vocals from song files more cleanly than an old low-bitrate MP3.
- Clean both ends. High-pass around 80–100 Hz for rumble and bass bleed; a de-esser for harsh esses and hi-hat ghosts.
- Mute the gaps. Faint leftovers are most audible between phrases — cut or gate them.
- For speech, denoise afterwards with your editor's noise reduction.
Recorded your cover? Hear it mastered.
Drop your finished mix into AI mastering and compare Warm, Balanced and Bright versions at matched loudness. Listening to the preview costs nothing; downloading the WAVs costs 10 credits or is included in a plan.
What is a vocal isolator?
A vocal isolator separates the voice in a recording from the music behind it. Ours uses an AI model on a cloud GPU to separate vocals from music and returns two files of the same length: the vocal on its own and the music without it.
How many credits does the vocal isolator cost?
Each song costs 4 credits (about $0.60 or less), and you see the price before you upload. Each split is a new GPU job, so running the same file again costs again; failed splits are refunded automatically. The trial pack is $2.99 for 20 credits, pack credits last 12 months, and plans from $7.99 a month include stem credits every month. See pricing.
Is a vocal isolator the same as a voice isolator for background noise?
Not quite. Many "voice isolator" tools clean up speech by removing hum, wind and room noise. This vocal isolator separates the voice from music, so it suits a song or narration over a music bed, but it won't clean up a noisy room recording.
How clean is the isolated vocal compared with LALAL.AI, Moises or a DAW?
On modern pop, hip-hop and EDM the result is usually very clean; dense mixes, long reverbs and live recordings leave more bleed with any tool. LALAL.AI and Moises offer more stem types; here every split has a fixed credit price, shown first, paid from a monthly plan or a credit pack.
Can it separate the lead vocal from the backing vocals?
Not in the 2-stem split: lead, harmonies and ad-libs all land in one vocal file. The 4-stem stem splitter (7 credits) can return backing vocals as a separate part, depending on the song.
Which formats can I upload, and what do I get back?
MP3 or WAV, up to 10 minutes per song. Both files download as MP3.
Does my file get uploaded?
Yes. Vocal separation needs a GPU, so your file is uploaded to our storage and processed by our separation provider. It is used only to create your two files, and download links stay valid for about a week.
Is it legal to isolate vocals from a song?
Your own recordings are fine; use the vocal isolator only on music you own or have permission to use. For anyone else's song, posting, releasing or selling anything built on the isolated vocal needs permission from the rights holders. This is general information, not legal advice.
Try the vocal isolator on your next song.
The voice on its own and the music without it — usually in a few minutes, priced before you upload.
Drop a file to start
MP3 or WAV · songs up to 10 min