Know what's real
before you publish it.

Tayanch checks images, video, and audio for signs of AI generation and returns a clear verdict — one API call, no guesswork.

Try the demo Read the docs ›
Authentic
Natural skin texture and light — the irregularity of the real.
AI-generated
Repeating micro-patterns the eye misses, flagged in the grid.
Images
Video
Audio
Instant verdicts

How it works.

01
Send data
Upload an image, video, or audio file to the /analyze endpoint.
02
Model scores it
Vision-transformer embeddings and physical-consistency checks combine into a single score.
03
Get a verdict
Receive a clear verdict and the confidence behind it — tuned for your false-positive tolerance.

Try a sample check.

Upload an image and Tayanch returns a live verdict. Video and audio are coming soon.

Drop an image here, or tap to choose a file.

A plain HTTP API.

Send a file, get a verdict and the confidence behind it back. Works for images, video, and audio.

# request
curl https://api.yourdomain.com/analyze \
  -H "X-API-Key: $KEY" \
  -F "image=@photo.jpg"
# response
{
  "verdict": "ai",
  "label": "Likely AI-generated",
  "p_ai": 0.98,
  "confidence": 96
}
Images, video, and audio
One endpoint family handles stills, clips, and audio — each response includes the verdict, a probability, and a confidence.
Calibrated thresholds
Verdicts are tuned per use case, so you can trade recall for a lower false-positive rate.
Simple integration
REST over HTTPS, JSON responses, and API-key auth via the X-API-Key header.

Pricing.

Monthly plans, or pay only for what you send through your API key.

Starter
$0
per month · 500 checks
For evaluating the API before integrating it.
Get started
Team
$249
per month · 50,000 checks
For products running detection in production.
Start trial
Enterprise
Custom
volume pricing
Dedicated thresholds, SLAs, and support.
Contact sales