Skip to main content

​
Other Content Types

ZebraTruth’s primary product is video compliance. Image + text endpoints are available for partial / text-only content reviews.
Most customers run video compliance checks. Image + text are listed here for completeness — see Quickstart for the video flow.

​
Image Compliance

Endpoint: POST https://api.zebratruth.ai/v1/compliance/check-image Single-image compliance check. Sync (returns the full report in the same request). Multipart form upload — no separate upload-URL step. 
curl -X POST https://api.zebratruth.ai/v1/compliance/check-image \
  -H "Authorization: Bearer $ZEBRATRUTH_API_KEY" \
  -F "image=@your-ad.jpg" \
  -F "jurisdictions=us,eu" \
  -F "platforms=instagram,facebook"
Returns the same report envelope shape as video, with one key difference: image checks don’t carry timelineLocation (no time dimension) — they carry bbox for spatial regions on the image. Detects: brand logos, watermarks, restricted categories (alcohol/tobacco/firearms/etc.), content safety (adult/violence/racy), and C2PA metadata. Use video instead if: your content is a video. Submitting individual frames to the image endpoint costs more, loses temporal context (no transcript), and misses claims that only appear when frames are read together.

​
Text Compliance

Endpoint: POST https://api.zebratruth.ai/v1/compliance/check Sync ad-copy / script / caption compliance check. JSON body, ≤500 words. 
curl -X POST https://api.zebratruth.ai/v1/compliance/check \
  -H "Authorization: Bearer $ZEBRATRUTH_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Try our miracle weight-loss pill — guaranteed results in 30 days!",
    "jurisdictions": ["us"],
    "platforms": ["tiktok"],
    "mode": "fast"
  }'
Detects: unsubstantiated claims, missing disclaimers, restricted-category language, regulatory compliance (FTC §5, EU UCPD, UK ASA), platform AI-disclosure language. Use video instead if: your content has spoken claims or on-screen text — the video pipeline auto-extracts these via transcription + per-frame OCR and analyzes them in temporal context.

​
When to use which

You haveUse
A video ad (mp4/webm/quicktime)/v1/compliance/check-video ← primary product
A single image (jpg/png/gif/webp)/v1/compliance/check-image
Ad copy / script / caption only/v1/compliance/check
The video endpoint dominates because it captures all signals in one call: spoken + on-screen + visual.

​
API reference

Full request + response schemas for all three endpoints:

OpenAPI Reference

Endpoint specs, request bodies, response envelopes, error codes.