AI Deepfake Detector
Smartly detect deepfake content in images, videos and audio, including face swap recognition, voice clone detection and AI-generated content analysis
Detector Interface
Interactive detector will be available soon
Features
- ✓ Detects AI face swaps (Deepfake), face synthesis and expression reenactment videos
- ✓ Authenticates voice clones and AI-synthesized audio
- ✓ Multi-modal deep learning forgery analysis that identifies boundary blur, lighting anomalies and other traces
- ✓ Batch detection — analyze multiple images and videos at once
- ✓ Results include confidence scores and detailed forgery region annotations
How to Use
- Upload the image, video or audio file you want to check
- Choose detection type (face swap / voice clone / all types)
- Click start detection and wait for AI analysis to finish
- Review the results, confidence scores and forgery region annotations
FAQ
What is AI Deepfake Detector?
An online AI deepfake detection tool. It smartly detects deepfake content in images, videos and audio, including face swap recognition, voice clone detection and AI-generated content analysis. Input: the image, video or audio file you need to check. Output: detection results, confidence scores and forgery region annotations with one-click copy and download. Everything runs in your browser — no uploads.
How accurate is the detection?
The detector is built on multi-modal deep learning models and achieves over 95% accuracy on public benchmark datasets for common Deepfake generation methods. Real-world results can vary with video compression and resolution, so we recommend combining it with human judgment.
Which file formats are supported?
Images: JPG, PNG, WEBP. Videos: MP4, MOV, WebM. Audio: MP3, WAV, M4A. Maximum file size is 200MB per file.
What types of forgery can it detect?
It covers mainstream deepfake types: AI face swaps, expression reenactment, face synthesis, voice cloning, AI dubbing, lip-sync forgery and a subset of AI-generated images.
How long does detection take?
A single image typically completes in 5-15 seconds. Videos are processed by duration, usually within 3 minutes. Batch jobs are queued in the background.
Can I save the detection results?
Yes. Reports can be exported as PDF or CSV, including confidence scores, forgery region screenshots and analysis notes for evidence and archiving.