Music Analysis API: The TrackScore Developer API Is Here

A music analysis API turns a track into structured data: software submits audio and gets an objective verdict back as JSON. The TrackScore.AI™ Developer API returns a 0-100 score, a 13-grade letter (A+ to F), genre fit, frequency balance, dynamics, stereo metrics, and the Klaus™ diagnosis, a written assessment with an improvement plan. It is in early access today.

Why a Music Analysis API

Since launch, TrackScore has scored thousands of tracks for producers through the web app: upload a mix, get a verdict in about a minute. But some of the most interesting requests came from people who did not want a web page. Distributors screening submissions, labels triaging demo inboxes, sample marketplaces scoring catalogs, and producer tools that want mix feedback inside their own UI all need the same thing: the analysis as data, not as a dashboard.

That is what the Developer API is. Every number TrackScore computes, from the 7-band frequency balance to the 13-grade letter, plus the full Klaus™ written diagnosis, comes back in a single structured response you can store, rank, filter, or render however you like.

What the API Returns

One completed analysis is one JSON object. The response includes the overall score and grade, five pillar scores, tempo and key detection (with Camelot notation for harmonic mixing), genre detection against 9 electronic-genre profiles, frequency balance with per-genre ideals, loudness and dynamics, stereo measurements, arrangement sections, prioritized feedback, and the diagnosis:

{
  "object": "analysis",
  "id": "9b2f6c1e-...",
  "status": "complete",
  "diagnosis_status": "ready",
  "score": 87,
  "grade": { "letter": "B+", "label": "Very Good" },
  "scores": { "danceability": 91, "mix_quality": 88, ... },
  "tempo": { "bpm": 124.0, "regularity": 0.94 },
  "key": { "key": "A", "scale": "minor", "camelot": "8A" },
  "genre": { "detected": "deep_house", "name": "Deep House" },
  "frequency": { "bands": { ... }, "ideal": { ... }, "score": 84 },
  "mix": { "loudness_lufs": -8.2, "dynamic_range_db": 7.9, ... },
  "diagnosis": { "status": "ready", "summary": "The low end is dialed in. ..." },
  "feedback": [ { "type": "warn", "category": "frequency", "text": "..." } ]
}

Scores land first, and the full written diagnosis fills in shortly after; a diagnosis_status field tells you which phase you are looking at. A typical track completes in about a minute.

How It Works: Two Requests

1. Create the analysis. POST /api/v1/analyses with the file metadata and an idempotency key. You get a 202 with the analysis id and a short-lived upload target.
2. Send the bytes. POST the raw audio to the returned upload URL. WAV, MP3, AIFF, and FLAC are accepted, up to 200 MB and 20 minutes per file.

Then either poll GET /api/v1/analyses/:id until status is complete, or skip polling entirely and register a webhook endpoint. Every delivery is signed with an HMAC-SHA256 TrackScore-Signature header so your server can verify it came from us:

Zero-Storage, Now as a Contract

The API inherits the same zero-storage architecture as the web app: audio is processed in memory, analyzed, and discarded. No disk, no object storage, no queue, and no AI training on your audio, ever. For an API this matters even more, because integrators are often submitting unreleased catalogs at scale on behalf of their own users.

So we wrote it into the legal terms, not just the architecture docs: zero-storage audio handling is section 1 of the API Terms of Service, as a binding commitment.

Build Against It for Free

Keys prefixed ts_test_ return a deterministic sample analysis with the exact production response shape. Test calls never touch the analysis engine and never spend a credit, so you can wire up submission, polling, webhook verification, and rendering end to end before your first live call.

Pricing

API access is included with Pro & Studio and uses your monthly credits: one analysis costs one credit, the same as the web app. Test mode is free. Building at volume or integrating a platform? Contact us for partner pricing.

What You Can Build With It

• Demo triage for labels: score every inbound demo and surface the strongest submissions first, with the Klaus™ diagnosis attached as a first-pass review.
• Distribution quality gates: catch clipping, loudness, and frequency-balance problems before a release goes out the door.
• Catalog intelligence: batch-score a sample library or back catalog and make it searchable by tempo, key, genre fit, and mix quality.
• Producer tools: give your users mix feedback inside your own product, with scores you can render any way you like, or drop in the embeddable score badge.

Early Access: What Is Live Today

The developer documentation is live now: a five-minute quickstart, an interactive OpenAPI 3.1 reference with cURL, Node.js, and Python examples, complete error-code documentation, and a published versioning policy with a 12-month sunset commitment for any breaking change. Keys are rolling out to Pro & Studio accounts through the Developer console, and official Node.js and Python SDKs generated from the OpenAPI spec are on the way.

If you are building something that needs to understand how a track actually sounds, I would genuinely love to hear about it: hello@trackscore.ai.

Music Analysis API: FAQ