Solmi's AI acapella extractor pulls the vocals out of any song and gives you a clean, isolated acapella in seconds. Upload an audio file, and the AI separates the vocal layer from the instrumental — download the vocals-only track for remixes, mashups, DJ edits, or sampling. No software to install: sign in with a free account, and each song uses 10 credits, with free credits refilling to 10 every day.
1. Sign in with a free account. 2. Upload an MP3, WAV, M4A, AAC, OGG, or FLAC file up to 70 MB, or paste a song link, and pick MP3 or WAV for the download. 3. Start the split: the AI separates the vocals from the music and uses 10 credits. 4. Play the acapella and the instrumental on the page, then download the acapella, and the instrumental too if you want it. If a split fails, the credits go back to your account automatically.
One run splits the whole song into two tracks. Everything that is sung lands in the acapella: the lead vocal, backing vocals, harmonies, and ad-libs. Everything else, from drums and bass to guitars, keys, and synths, stays in the instrumental.
| Spec | Detail |
|---|---|
| Input | MP3, WAV, M4A, AAC, OGG, or FLAC up to 70 MB, or a song link |
| Output | Two tracks: the acapella (vocals only) and the instrumental |
| Download format | MP3 or WAV |
| Price | 10 credits a song; free credits refill to 10 every day |
| Failed split | Credits refunded automatically |
| Video files | Not accepted: extract the audio as an MP3 first |
Start from the best copy of the song you have: a WAV, a FLAC, or a high-bitrate MP3 separates more cleanly than a file that has been compressed several times or re-recorded from a speaker. Studio mixes with a clear, centered vocal come out cleanest. Reverb and delay on the voice are part of the vocal, so they come out in the acapella too, and on live recordings the crowd can bleed into the vocal track.
All three run the same separation model. Use the acapella extractor when the voice is what you are after: mashups, vocal chops, DJ edits, or studying a singer's phrasing. Use the vocal remover when you want the instrumental for karaoke or a cover; it gives you both tracks as well. Use the stem splitter when you also need the drums, bass, and other instruments on their own tracks, with 4 stems for 15 credits or 6 for 20.
Yes, with a free account and no credit card. Each song uses 10 credits, and free credits refill to 10 every day, enough for one song a day. The AI outputs an isolated vocals-only track plus the matching instrumental.
Upload the song (MP3, WAV, M4A, AAC, OGG, or FLAC) and the AI separates the vocals from the music in seconds. Download the acapella as a standard audio file ready for any DAW.
The extractor uses studio-grade AI source separation, so on most modern mixes the acapella comes out clean enough for remixing and sampling. Dense mixes with heavy vocal effects can leave slight artifacts.
Yes — every extraction outputs both stems. If it's the instrumental you're after, the same tool doubles as a vocal remover.
Upload an audio file up to 70 MB (MP3, WAV, M4A, AAC, OGG, or FLAC), or paste a song link. You get the acapella and the instrumental, each downloadable as MP3 or WAV.
No. The AI splits the song into two stems, vocals and instrumental, so the lead, backing vocals, and harmonies all land together in the acapella. To pull drums, bass, and other instruments apart as well, use the stem splitter.
Yes. The model separates the voice by its sound, not by the words, so rap, sung vocals, and any language work the same way. The mix matters more than the genre: a dense, heavily processed mix can leave faint traces of instruments in the acapella.
Extracting the vocal does not change who owns the recording: to publish or sell a remix built on someone else's vocal, you need permission from the rights holders. Acapellas from your own songs are yours to use however you like.