AI-Generated Image Heuristic Detector
Visual + metadata flags — Midjourney, DALL-E, SDXL
Tick what applies to the image:
📚 Learn more — how it works, FAQ & guide Click to expand
Learn more — how it works, FAQ & guide
Click to expand
AI-Generated Image Heuristic Detector
Visual + metadata flags for AI-generated images. Hand artifacts, text rendering, eye symmetry, EXIF model name, C2PA tag. Manual sniff-test, not court evidence.
How to use this tool
- 1
Tick visual flags
Hand artifacts, eye symmetry, etc.
- 2
Check metadata
EXIF, C2PA, model name in filename.
- 3
See probability
AI-generated likelihood + confidence.
Frequently Asked Questions
How accurate is heuristic detection?
Modern Midjourney v6 / SDXL / DALL-E 3 are nearly indistinguishable from photos for skilled prompts. Heuristics catch ~70% of casual AI generations. Better: C2PA cryptographic provenance.
What is C2PA?
Coalition for Content Provenance and Authenticity. Cryptographic metadata signed at creation. Adobe, OpenAI, Microsoft support. Future-proof verification — when implemented.
Hand + text rendering tells?
Top historical tells. v6+ models fixed hands mostly; text still bad. 2026 generators sometimes nail both. Don't rely on these alone.
You might also like
🔒
100% Privacy. This tool runs entirely in your browser. Your data is never uploaded to any server.