Genuine image
No detected AI generation or editing.
Developer API · Available now
AI Image Detector API
Upload an image and receive a real, fake or inpainting verdict with confidence and per-class probabilities for review tools, publishing systems and content platforms.
- JPEG, PNG, WebP + more
- 10 MB maximum
- Multipart upload
Prediction
REAL
Confidence 99.64%
Three-class response
Read the verdict, confidence and class probabilities.
The response identifies the predicted class and returns its confidence alongside probabilities for real, fake and inpainting.
{
"prediction": "real",
"confidence": 0.9964,
"probabilities": {
"real": 0.9964,
"fake": 0.0013,
"inpainting": 0.0022
},
"credits_charged": 6
}
Fully generated
Image is classified as AI-generated.
Partially edited
Authentic image with AI-edited regions.
Image request
Send one image and receive the full result.
Use an API key with the image detector scope and send the file as multipart form data. Store the verdict, confidence and class probabilities with your review record.
Open implementation details →/api/image-detector/predict/
curl -X POST \
https://developer-portal.proofademic.ai/
api/image-detector/predict/ \
-H "X-API-Key: YOUR_API_KEY" \
-F "[email protected];type=image/jpeg"
Operational details
Validate image files before upload.
Accept supported file types up to 10 MB, send the image as multipart form data and use the health endpoint to check service availability.
Image upload
- Formats
- JPEG, PNG, WebP, GIF, BMP, TIFF
- Maximum
- 10 MB per file
- Body
- multipart/form-data
Verdict plus probabilities
Each completed response returns a verdict, overall confidence and probabilities for every supported class.
Photo Real
Generated Fake
Edited Inpainting
Output Confidence
Service health
GET /api/image-detector/health/
healthy
Circuit closed · engine ready
Health response →Use cases
Use AI image detection in content review.
01 / Education
Submission review
Flag images that may need closer review in an academic integrity process.
For institutions ↗ 02 / PublishingEditorial screening
Check photographs and submitted assets before publication or licensing.
For publishers ↗ 03 / PlatformsContent review
Use the returned prediction and probabilities to route images for human review.
Explore solutions ↗API questions
Before your first request.
The prediction is real, fake or inpainting. The response also includes overall confidence and a probability for every class.
Send your API key in the X-API-Key header. The key must include the image_detector scope.
The endpoint accepts JPEG, PNG, WebP, GIF, BMP and TIFF files up to 10 MB as multipart form data.
The response includes a real, fake or inpainting verdict, overall confidence and a probability for every supported class.
Credits are reserved before analysis and automatically refunded when detection fails.
Developer resources
Find request formats, response fields and errors.
Manage keys and usage in the platform. Use the API reference for current request formats, response fields, limits and errors.
Research and guidance
Learn how to review AI image detection results.
Ethics · May 2025
Understanding AI Detection Technology: How It Works and Why It Matters
As AI writing tools become more sophisticated, the need for reliable detection methods grows. Learn how modern AI detection works and why…
Read the article →
AI Content Detection · Dec 2025
Understanding False Positives in AI Detection: A Comprehensive Guide for Educators and Content Creators
Understand false positives in AI detectors. Learn why they occur, compare accuracy rates across platforms, and discover best practices to minimize false…
Read the article →
AI Content Detection · Jun 2026
Best AI Detector in 2026: 10 Top-Rated Tools Ranked
Proofademic is the best AI detector in 2026. We ranked 10 tools on accuracy, paraphrase resistance, and academic fit - here's the…
Read the article →Start building
Bring image detection into your workflow.
Create an API key, upload an image and receive the verdict, confidence and all three class probabilities.