Release is submitted
Your system sends each track's audio URL and duration as it lands, before QC and delivery.
Pre-release screening for distributors & labels
Catch AI-generated tracks before they reach streaming services
Screen every delivery for Suno, Udio and other music generators. Each track gets an AI probability, the likely generator and a section-by-section timeline — over the API or in the browser.
Pay per audio minute Failed checks refunded Deleted within 24 hours
track07_final_master.wavSample report
96%
AI-generated
The audio carries strong generator fingerprints across the track.
Likely source
Most likely made withSuno
Timeline
0:003:04
AI signalMixedHuman signal
Stems
Vocals97.8% AI · Suno
Instrumental88.5% AI · Suno
Detection is probabilistic. Use it to route tracks to review, not as the sole basis for takedowns. How to read results
Where screening fits in your release pipeline
A free "paste a song" page can't screen a delivery queue. The API is built for that: one call per track, a stable JSON result, and bands you can map to your own policy.
Detection runs
Usually under a minute per track. Poll for the result or check it later from the dashboard.
You decide
Auto-approve human reads, queue inconclusive ones for a listener, apply your AI policy to the rest.
One endpoint. Priced by the minute.
1 credit covers up to 5 minutes of audio. API packs start at $19 for 1,000 credits — about $0.019 per track — and never expire.
- Bearer-key auth, keys managed in the dashboard.
- Same report structure in the API and the web app.
- Credits are refunded automatically when a check fails.
- Audio is never used for training and is deleted within 24 hours.
curl -X POST https://musicdetector.org/api/v1/detections \
-H "Authorization: Bearer $AIMD_API_KEY" \
-H "Content-Type: application/json" \
-d '{"audio_url":"https://cdn.example.com/release/track01.wav",
"duration_seconds":214, "file_name":"track01.wav"}'
# → { "data": { "id": "tsk-…", "status": "processing", "credits": 1 } }Built for release teams
11 generators scored
Suno, Udio, Mureka, ElevenLabs Music and more. "Source unclear" when no single generator dominates.
Section-level evidence
A timeline shows whether AI signal runs through the track or sits in one section.
Vocal and backing reads
Separate scores reveal AI vocals on a human production, or the reverse.
Private by default
Unreleased masters are deleted within 24 hours and never used to train anything.
Honest confidence bands
Mixed evidence is reported as Inconclusive, so you know which tracks need a listener.
Audit-ready JSON
Keep the report per track alongside the release for disputes and partner questions.
We publish how often it's right
"Does an AI music detector actually work?" is a fair question. Our accuracy page explains how results are scored, where the detector struggles, and our benchmark across generators, versions and re-encoded files.
Accuracy & limitsHarder cases we test for
- New generator versions released after the model was trained
- Low-bitrate MP3 and streaming-quality re-encodes
- Pitch-shifted, time-stretched and trimmed files
- AI stems mixed with live recordings
Pricing
Free
$0 · 3 checks every day with a free account.
Creator
$9.90/month · 200 checks for artists, A&R and curators.
API
From $19 · billed by audio length, credits never expire.
Questions from release teams
Call POST /api/v1/detections with a signed https URL to the audio and its duration when a release is submitted, then poll the detection until it completes. Route each track by its verdict band: approve, send to human review, or apply your AI policy.
Screen your next delivery
Create an API key in the dashboard and send your first track in minutes.