AUDIO MODEL · META
Demucs
Demucs separates a recording into audio stems such as vocals, drums and bass. It is useful for isolating parts of a track for editing or analysis, with the available stem choices determined by the selected model settings.
What you can do
- Audio editing
Sources
Pricing
Renderly shows the exact price for your selected settings before you confirm. Prices are in USD and credits; one credit has a balance value of $0.01.
Use it through the API
Request a price quote with this model’s ID. Quoting does not spend credits; review the amount, then submit the job with the quote ID.
cURL · price quote
# Set RENDERLY_API_URL and RENDERLY_API_KEY as shown in the API guide.
curl --fail-with-body "$RENDERLY_API_URL/quotes" \
-H "Authorization: Bearer $RENDERLY_API_KEY" \
-H "Content-Type: application/json" \
--data '{"modelId":"model-50c0087f064236a8847b34327ca90d29","operation":"audio-to-audio","sku":"selected-settings-v1","input":{"parameters":{"p0":["vocals","drums","bass","other","guitar","piano"],"p1":1,"p2":0.25,"p3":"YOUR_UPLOADED_ASSET_ID","p4":"mp3"}}}'SDKs
npm install renderly-ai
pip install renderly-ai