Segmentation maps a song's arrangement. The job returns the sections in order, each with a start and end time in seconds and a label such as intro, verse or chorus.
Every section also carries the features you ask for: energy, loudness, danceability and tatum among them. That makes it straightforward to find the drop, pull a chorus out for an edit, or line two tracks up by arrangement instead of by ear.
Send the mix on its own, or send stems alongside it to analyse components separately (bass, beat, melody, vocal). A genre hint sharpens the boundaries. Call it at /functions/run/segmentation, from audial.segment in Python, or as the segment MCP tool.
What you get
- Section list with start and end timestamps in seconds
- Labels such as intro, verse and chorus
- Per-section features: energy, loudness, danceability and tatum
- Optional per-component analysis for bass, beat, melody and vocal
- Genre hint to improve section boundaries
- Accepts stems from the stem splitter for a finer read
Use it from code
curl -X POST "https://api.audialmusic.ai/api/functions/run/segmentation" \
-H "X-API-Key: your_api_key" -H "X-User-ID: your_user_id" \
-F "userId=your_user_id" \
-F "original=@/path/to/track.mp3" \
-F 'segmentationParameters={"components":["bass","beat","melody","vocal"],"analysisType":"select_features","featuresToExtract":["energy","loudness"],"genre":"Tech House"}'import audial
audial.config.set_api_key("your_api_key")
audial.config.set_user_id("your_user_id")
segments = audial.segment(
"track.mp3",
components=["bass", "beat", "melody", "vocal"],
analysis_type="select_features",
features=["energy", "tempo", "loudness", "danceability"],
genre="electronic",
)
print(segments["execution"]) # sections with start, end and type# In Claude, Cursor or any MCP client with audial-mcp installed:
"Segment ~/Music/track.mp3 and tell me the timestamp where the first chorus starts."Frequently asked questions
A list of sections in order. Each one has a start time, an end time, a type such as intro, verse or chorus, and whichever features you requested.
Yes. Sections are labelled by type and carry energy and loudness values, so the loudest chorus section is usually the drop you are looking for.
Energy, loudness, danceability and tatum are the common ones. Ask for the set you need with the features list and the response carries them on every section.
It can. Pass stem URLs alongside the mix and the job analyses the bass, beat, melody and vocal components separately, which sharpens the component read.
Yes. Post to https://api.audialmusic.ai/api/functions/run/segmentation with your X-API-Key and X-User-ID headers. An active subscription is required, as with every Audial function.