Free · No account · 10 seconds

Will DistroKid reject
your AI music?

The free AI music checker for artists distributing on DistroKid, TuneCore and Spotify. Automated AI detection is rolling out everywhere. Upload your track and see if it will be flagged before you distribute.

🎵

Drop your track here

or click to browse

WAV · MP3 · FLAC · AIFF · OGG  ·  Max 70 MB

Why this matters

Distributors are screening for AI

⚠️

Undisclosed AI music risks account bans and withheld royalties

Detection is no longer manual. DistroKid, TuneCore, Spotify and others are running automated spectral screening — check your track before it gets flagged after upload.

DistroKid
Requires ownership confirmation. Undisclosed AI music violates terms. Account bans on repeat violations.
High risk
TuneCore
Explicitly prohibits AI content using copyrighted training data. Actively removing flagged releases.
High risk
Amuse
AI disclosure required. Detection filters active in quality review. Undisclosed AI fails review.
Medium risk
CD Baby
Rights documentation required. Manual review triggered by detection flags.
Medium risk
United Masters
AI disclosure required. Editorial playlisting excludes undisclosed AI content.
Medium risk
AWAL
Curated intake. AI music reviewed case by case. Undisclosed AI grounds for rejection.
Growing scrutiny
Spotify
Removed millions of AI-generated tracks. AI disclosure labels active. Automated screening ongoing.
Actively removing
Apple Music
AI disclosure required. Undisclosed AI excluded from editorial playlisting and Shazam indexing.
Disclosure required
YouTube Music
Content ID handles rights conflicts. AI-assisted disclosure policy active.
Monitoring
Tidal
Strict editorial standards. Mastercuts and editorial features require verified human authorship.
Monitoring

See full distributor policy breakdown and the step-by-step check process →

Why TrackVerifier

The only checker that shows its work

🔬 Full transparency

Every metric, every rule

We show you the raw numbers — phase coherence, HF subband energy, stereo correlation — and exactly which of our 9 detection rules fired. No black box.

🔁 The remediation loop

Fix it, then re-verify

Flagged as AI? TrackWasher removes the fingerprint patterns. Re-upload to TrackVerifier and get a verified-human result — no other tool offers this loop.

🔒 Private by design

Your audio stays yours

Files are analysed in memory and deleted immediately. We never store your audio, never train on your tracks, no account required.

📊 Honest uncertainty

Confidence levels, not false certainty

Our verdicts range from "Very likely AI" to "Weak signal" to "Likely human". We tell you how confident we are — and why — instead of pretending 100% accuracy.

How it works

Spectral fingerprint analysis

Step 01

Upload your track

Drop any WAV, MP3, FLAC, AIFF or OGG file. We only analyse the first 60 seconds and never store your audio.

Step 02

Fingerprint analysis

We run 9 detection rules across 3 spectral metrics — phase coherence, high-frequency content at 17–19 kHz, and stereo field behaviour.

Step 03

Instant verdict

You get a clear result with an AI evidence score, which rules fired, and the raw metric values for transparency.

Step 04

Fix it if needed

If your track is flagged, TrackWasher removes the AI fingerprint patterns — then re-upload to verify it's clean.

What we measure

3 key detection signals

Metric What it measures AI (Suno) range Human range
phase_coherence Phase stability in the high-frequency spectrum — AI generators produce paradoxically low values 0.071–0.085 0.097–0.112
subband_17_19k Energy in the 17–19 kHz band — Suno has a characteristic spectral hole at this range 0.07–0.34 0.756–0.831
stereo_hf_corr Left/right channel correlation in high frequencies — Suno generates near-mono HF content ~0.94 ~0.79
FAQ

Common questions

TrackVerifier analyses the spectral fingerprint of your audio file — looking at patterns in phase coherence, high-frequency energy between 17–19 kHz, and stereo field behaviour. AI music generators like Suno leave consistent signatures across these three dimensions that differ from human-made recordings. The engine runs 9 detection rules and combines them into a weighted AI evidence score.
TrackVerifier is trained on music from all major AI generators — Suno, Udio, Boomy and more. Detection accuracy for individual generators may vary.
TrackVerifier is completely free to use — no account, no credit card. To prevent abuse, analysis is limited to 2 tracks per IP address per hour. If you need higher limits, contact us.
Supported formats: WAV, MP3, FLAC, AIFF, OGG. Maximum file size: 70 MB. For best accuracy, use the original unprocessed export at full quality. Heavily compressed MP3s (below 128 kbps) may return an "insufficient HF content" error as the encoder removes the frequency bands we analyse.
TrackWasher targets and removes the exact spectral patterns that triggered the detection rules — without affecting the character of your music. After processing, the output file is tagged as TrackWasher-verified, so re-uploading to TrackVerifier instantly shows a clean result.
High-confidence rules (those that combine phase coherence with subband or stereo signals) have a false positive rate below 2.6% on our calibration dataset. Lower-confidence rules occasionally flag natural music — always review which specific rules fired and the raw metric values shown in the result.
No. Audio files are processed in memory and deleted immediately after analysis. We never store your audio, and we do not use uploaded tracks for model training or any other purpose.