Diffusion Artifacts
Detect subtle diffusion patterns, noise inconsistencies, and generation artifacts from Stable Diffusion, DALL-E, and Midjourney
Free online image check
Upload a JPG or PNG to check for AI generation, diffusion and GAN artifacts, face swaps, and image manipulation. Review a probability score with supporting evidence.
Comprehensive AI image and manipulation detection
Detect subtle diffusion patterns, noise inconsistencies, and generation artifacts from Stable Diffusion, DALL-E, and Midjourney
Identify replaced faces with lighting inconsistencies, boundary artifacts, and forensic analysis
Detect GAN-generated images from StyleGAN, BigGAN, and other generative models
Identify fully AI-generated images across multiple styles and generation techniques
Detect edited regions with inpainting tools and localized AI modifications
Identify unnatural geometric artifacts common in AI-generated images
Understand exactly why content was flagged
Understanding model constraints for optimal results
Highly compressed images, heavy filters, or artistic edits may increase false positives. Use confidence bands for thresholds.
Sophisticated anti-detection techniques may evade detection. Re-test against emerging attacks and use source context with the result.
Performance varies across novel generators and styles. Representative evaluation should include unseen models, edits, and compressed samples.
Get started in minutes with our unified API
curl -X POST https://app.zerotrue.app/api/v1/check \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"input": {
"type": "image",
"value": "https://example.com/image.jpg"
},
"is_private_scan": true,
"is_deep_scan": false,
"idempotency_key": "req-123"
}'The free web checker accepts JPEG and PNG files up to 10 MB. Confirm current API formats and limits for production workflows.
Processing time depends on image dimensions, format, input method, and scan mode.
No detector covers every generator or transformation. Coverage changes as image models and editing workflows evolve.
Face swap results are probabilistic. Review confidence, localized evidence, source context, and limitations before making a decision.
Retention can depend on the plan and scan mode. Confirm current retention and deletion requirements before submitting sensitive images.