What is Suno Extract Vocals?
Suno Extract Vocals isolates the vocal line from a section of a Suno clip. It works on a clip_id returned by an earlier Suno task, so the song has to exist in Suno already. You set the section with vocal_end_s, which is required, and vocal_start_s, which defaults to 0. The finished task returns a vox_audio_id for the isolated vocal, along with the source_clip_id and the start and end times you asked for, so you can match each result to its source. The time range is what sets it apart from Stems: rather than splitting the whole song into tracks, you pull out the singing from the exact passage you care about, such as one chorus.
Key features of Suno Extract Vocals
Three request fields, and one of them is optional.
Time-range selection
Set vocal_start_s and vocal_end_s in seconds to isolate the vocal from one verse, one chorus or one hook instead of the whole song.
Works on existing clips
Pass the clip_id of any Suno clip. There is no upload step, and the source clip is not changed.
Traceable results
Each result returns vox_audio_id, source_clip_id and the range you sent, so a batch of extractions stays easy to map back to its songs.
Sensible default start
vocal_start_s defaults to 0. For a vocal that starts at the top of the clip, you only need to send the end time.
Best for
Isolating a hook
Pull the sung hook out of a chorus to reuse in a remix, a jingle or a short video.
Studying a vocal line
Hear the melody and phrasing of one section without the instruments on top.
Remix preparation
Hand producers the isolated vocal of the section they want to rework.
Checking vocals before release
Listen to the isolated vocal of a key passage to catch pronunciation or timing problems that the full mix hides.
Batch processing many songs
Run the same range across a catalog of clips and track every result through source_clip_id.
Use cases
A remix app can let users drag a selection on a song's waveform, send the start and end of that selection as vocal_start_s and vocal_end_s, and drop the isolated vocal into a new project. A vocal coaching tool can isolate the chorus of a generated song so students hear the target melody on its own. When you need every part of the song separated rather than a vocal section, the Suno Stems API splits the whole clip into tracks.
API examples
Full parameter docs are at docs.unifically.com/models/audio/suno/tools.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "suno-ai/vox",
"input": {
"clip_id": "1901ec96-72c6-4c26-8e56-91028ca6070d",
"vocal_start_s": 45,
"vocal_end_s": 74
}
}'
The request returns a task_id. Poll GET /v1/tasks/<task_id> for the vox_audio_id, or include a callback_url so the result is posted to your server.
FAQs
People also ask
It isolates the vocals from part of a Suno clip. You send the clip_id and the time range in seconds, and the job separates the singing in that range from the rest of the mix.
Four fields. vox_audio_id identifies the isolated vocal, source_clip_id is the clip it came from, and vocal_start_s and vocal_end_s repeat the range you asked for.
clip_id and vocal_end_s. vocal_start_s is optional and defaults to 0, so leaving it out takes the vocal from the start of the clip up to vocal_end_s.
Extract Vocals works on a time range you choose and returns the isolated vocal for that range. Stems works on the whole clip and returns separate tracks, either vocals plus instrumental or a full set of instrument stems.
A passage where the vocal is clear and continuous, such as a verse or a chorus. Sections with long instrumental breaks give you less vocal to work with.
It is currently not billed. Check the pricing page for the latest rate.
