Extract the acapella from a song
Just the voice from a finished song, separated by an AI model in this tab. For remixes, mashups and learning a vocal part.
- Nothing is uploaded
- Works offline after first use
- Open source
- No watermark · limit is your device’s memory; phones hold less
Isolate the vocals of a song as an acapella for remixes, samples, transcribing lyrics or learning harmonies. It is free, and your files are not uploaded: everything runs in your browser on your device. The first run downloads the separation model (about 67 MB) and the engine that runs it to your browser, and after that the tool works offline; choosing MP3 output fetches a small encoder the first time you use it.
Done
Open the network tab. Your files never appear in it.
Add files and run the tool to see every request this tab makes.
Your files are processed on this device and never sent to a server. Reload the page to clear everything from memory.
An acapella is the vocal of a song on its own. Producers use them to remix and mash up tracks, singers use them to hear exactly how a phrase is sung, and they make lyrics far easier to write down. Official acapellas are rarely released, so the usual way to get one is to separate it from the finished song.
A worked example: a 4 minute FLAC of a song you want to remix at a different tempo. Drop it here; vocals only and WAV are already chosen, since a remix needs every bit of quality, and the result is a 16-bit stereo WAV of about 42 MB. The model predicts the music, and the vocal is what is left when that is subtracted from the song, so it lines up sample for sample with the original and drops into your project at bar one.
Expect some bleed. Hi-hats, snare hits and sustained synths that share the voice’s frequencies can leave traces in the acapella, most noticeable in gaps between phrases. A gate or a little editing in your audio editor cleans those up; for sampling a single phrase, pick one where the arrangement is sparse.
How it works
- Drop the song; vocals only and WAV are already chosen.
- Choose Both if you also want the instrumental, lined up with the vocal.
- Press Extract vocals. The first run downloads the model and its engine once, then the acapella downloads.
In your browser, not on a server
Your browser’s own decoders read the song, from an audio file or a video’s soundtrack, at 44.1 kHz stereo. It is cut into pieces of about six seconds, and each piece becomes a spectrogram (a short-time Fourier transform with a 6144-point window, as the model was trained with). UVR-MDX-NET Inst HQ 3, an MDX-Net model from the Ultimate Vocal Remover project (MIT), predicts the spectrogram of the music without the voice; it runs through onnxruntime-web (MIT) on the graphics chip with WebGPU where the browser has it, or on the CPU with WebAssembly. The answer is turned back into sound, and the vocals are the song minus the music.
Compared with upload-based tools: Online vocal removers upload the whole song, queue it, cap free minutes or songs per day, and ask for an account to download WAV. Here the song never leaves your device, the limit is your device’s memory (phones hold less), and the model works offline after the first run.
Verify it yourself. Give the page a short while after your first run to save itself for offline use, then turn off Wi-Fi or switch to airplane mode and run the tool again. Or open your browser's developer tools (F12, or Cmd-Option-I on a Mac), pick the Network tab, and watch it while you convert. The panel above does the same count for you.
Questions
Why WAV and not MP3?
The acapella will usually be edited, stretched or mixed again. WAV keeps it lossless so each step does not add compression artefacts; choose MP3 if you only want to listen.
Can it separate two singers?
No. Every voice ends up in the same vocal track, lead and harmonies together.
Can I publish a remix made with it?
Separating the vocal does not give you the rights to it. Remixes of other artists’ songs need permission to be released; for practice and private use there is no problem.
Is the song uploaded?
No. The song is decoded and separated in this tab, so unreleased tracks and demos never leave your device.
More video and audio tools
Vocal remover
Split a song into the music and the voice with an AI model running in this tab. Karaoke tracks and acapellas without uploading anything.
Remove background noise
Keep the voice, lose the hiss, hum and fan noise. Cleaned on this device, never uploaded.
Merge audio
Combine songs, voice notes or podcast segments into one file. Done on this device, never uploaded.
Mic test
Check that your microphone works before a Zoom, Teams or Discord call: speak and watch the level, then play back five seconds to hear how you sound. Nothing is uploaded.
Other conversions
Guides
- How to remove vocals from a songAI vocal removers pull the voice out of almost any song now, not just tracks with a lucky stereo mix. Here is how they work, and a free browser-based one.