de-artifact API

The same cleanup engine that powers the plug-in, available as a REST API for batch processing 100+ tracks at a time. Built for distributors, labels, mastering houses, and platforms verifying catalog-scale audio.

Pricing on request Contact for access Open Dashboard →

Throughput100+ tracks / batch
Latency~real-time per track
PresetsAll 7 plug-in presets
AuthAPI key (Bearer)

# Submit a batch — returns job ID
curl -X POST https://api.intrect.io/v1/cleanup \
  -H "Authorization: Bearer sk-int-..." \
  -H "Content-Type: application/json" \
  -d '{
    "preset": "ai_aggressive",
    "tracks": [
      {"id": "trk_001", "url": "https://..."},
      {"id": "trk_002", "url": "https://..."}
    ]
  }'

{
  "job_id": "cln_3f2a...",
  "status": "processing",
  "total": 2
}

# Poll results
curl https://api.intrect.io/v1/cleanup/cln_3f2a/results?format=jsonl
# → presigned URL per cleaned track + per-frame metadata

Built for catalog work

The plug-in is for the producer at the DAW. The API is for the company that processes thousands of tracks per week — automated ingestion, auditing, re-mastering, distribution-ready output.

Bulk submission

Submit up to 1,000 tracks per batch. Reference audio by R2 / S3 / HTTPS URL — no need to upload through the API itself.

Same 7 presets, same engine

HPSS focus, AI aggressive, AI vocal stem, lossy conservative, harmonic focus, balanced, q=0.6 20 dB. Switch per-track or per-batch.

Per-frame metadata

Cleaned audio + the model's mask predictions, GR per band, and HPSS split returned as compact JSONL — for forensic audit trails.

Webhook callbacks

Don't poll — register a webhook and receive a POST when each batch completes. Includes presigned URLs valid for 24 h.

SLA & private deployment

Enterprise tier: 99.9 % SLA, dedicated workers, optional on-premise deployment for catalogs that can't leave the org.

Same models as the plug-in

Bit-identical ONNX engine — what you hear in your DAW is what your batch will produce.

Use cases

Who	What they do with it
Music distributors	Pre-process AI-tagged uploads to catalog standard before delivery to streaming platforms.
Record labels	Restore separated stems delivered by producers — vocal stems from htdemucs, drum stems from AI tools.
Mastering houses	Codec-residual cleanup on lossy source masters before final EQ / loudness work.
Streaming platforms	Standardize uploaded AI-generated catalog at ingestion. Audit trail per track.
Sync / library	Clean AI demos for client preview before re-recording with humans.