Classify a song's music genre
Classify one song into BudgetPixel’s canonical music genre taxonomy — the same classifier that powers the site’s music gallery. Synchronous: the response carries the genre plus style/mood tags and a confidence score.
Input: one song (audio), max 20 MB and 7 minutes, as MP3, WAV,
OGG, M4A, or FLAC. Accepts a public URL, an uploaded-file URL from
POST /v1/uploads, a data URI, or
raw base64. Tip: a 7-minute WAV exceeds 20 MB — use MP3 for long tracks.
The classifier listens to a ~30-second excerpt sampled from ~25% into the
track (the clip field in the response tells you the exact window). Only
that excerpt is retained — privately, for ~1 day — never the full upload.
Pricing: flat 10 credits per song. Estimate via
POST /v1/cost with
{"model": "music-genre"}. Charged on success only — a classifier failure
costs nothing.
Authorizations
API key as a bearer token: Authorization: Bearer bpx_live_xxx
Body
One song (public URL, uploaded-file URL, data URI, or raw base64). Max 20 MB and 7 minutes. MP3/WAV/OGG/M4A/FLAC.
Response
The classification.
The excerpt the classifier heard.
Classifier confidence, 0-1.
Canonical genre, e.g. Electronic & Dance. One of the site's music genres.
URL-safe slug of the genre, e.g. electronic-and-dance.
Classification request id (audit reference).
One short mood phrase.
3-6 style/mood/instrument tags.