Tutorial · Tutorial · ~6 min
How to Extract Instrumentals from a Song
Karaoke backing tracks, remix stems, licensing reviews, practice tracks — an instrumental version of a song is one API call away. No DAW, no audio engineering, no manual EQ.
Step 1 — Understand what you'll get
'Instrumental' means the full mix minus the vocals. The model estimates the vocal stem and subtracts it, leaving everything else: drums, bass, guitars, keys. On clean sources this is indistinguishable from an official karaoke track to most listeners.
Step 2 — Submit the job
import requests, time
API = "https://api.mlslabs.io/v1"
HEADERS = {"X-API-Key": "YOUR_API_KEY"}
# sample audio included on the dashboard — or point at your own track
job = requests.post(
API + "/vocal-separator/jobs",
headers=HEADERS,
json={
"input": "s3://bucket/my-song.mp3",
"output": "s3://bucket/out/",
"mode": "music",
"tracks": ["vocals", "instrumental"],
},
).json()
job_id = job["job_id"]
print(job_id)Step 3 — Wait for the instrumental
while True:
st = requests.get(API + "/vocal-separator/jobs/" + job_id, headers=HEADERS).json()
if st["status"] == "done":
print(st["output"]["instrumental"])
break
if st["status"] == "failed":
raise SystemExit("failed — credits auto-refunded")
time.sleep(2)Step 4 — QC the result
- A/B against the original: vocals should be gone, groove intact
- Check for 'underwater' artifacts in the low mids (chorus/echo bleed)
- If bleed is audible, re-run with mode=music and a higher-quality source
- For karaoke use, confirm the lead vocal is fully gone in the first verse
FAQ
Common questions
Is the instrumental identical to an official karaoke version?
Extremely close on clean sources, but not byte-identical: expect a few artifacts you can hear on reference monitors. For practice, streaming and remixing it is excellent.
Can I get just the vocals too?
Yes — request both stems in the same job (tracks: ["vocals", "instrumental"]).
Does this work for any genre?
Yes — the model handles pop, rock, hip-hop, EDM and classical. Dense, heavily-processed mixes are the hardest.
Keep exploring
Related guides
Vocal Separator API
Isolate vocals or instrumentals from speech and music mixes.
Learn moreMake Karaoke Version In Python
Make a karaoke version of a song in Python: separate the vocal stem, keep the instrumental, export a clean backing...
Read guideAudio Source Separation Explained
Audio source separation explained: how modern models split music into vocals, drums, bass and other stems — spectro...
Read guide