How does ScanTheFake detect AI-generated content?

ScanTheFake uses multiple AI models to analyze content for authenticity. For images, we examine pixel patterns, metadata, and AI model fingerprints from DALL-E, Midjourney, Stable Diffusion, and Adobe Firefly. For text, we measure perplexity, burstiness, and stylistic markers to detect ChatGPT, Claude, and Gemini. For videos, we perform frame-by-frame temporal analysis and facial landmark detection to identify deepfakes from Sora, Runway, and Veo. For audio, we analyze acoustic patterns specific to Suno, Udio, and Riffusion.

How can I scan AI generated videos for deepfakes?

To scan AI generated videos, go to the Video Scanner page on ScanTheFake and upload your video file (MP4, WebM, or MOV, up to 3 minutes). Our AI performs frame-by-frame temporal analysis, examining facial landmarks, motion consistency, and AI model signatures. Results are returned in seconds with a confidence score showing the probability the video is AI-generated or a deepfake.

Is ScanTheFake free to use?

Yes, ScanTheFake offers a free plan that includes 3 image scans, 1 video scan, 1 audio scan, and 3 text scans per month. For more usage, Pro ($9.99/mo) includes 200+ scans per month, and Ultra ($24.99/mo) offers unlimited scanning across all content types.

Are my uploaded files stored?

No. Your files are processed in memory and immediately discarded after analysis. We never permanently store your uploaded content. We only retain anonymized scan metadata (file type, size, verdict) for aggregate statistics, which contains no personally identifiable information.

What types of content can ScanTheFake detect?

ScanTheFake can analyze five types of content: images (JPEG, PNG, WebP, GIF), videos (MP4, WebM, MOV), audio/music (MP3, WAV, OGG, FLAC, M4A), text (paste directly or upload PDF, Word, or TXT files), and links (direct URLs or social media posts from Twitter/X, Instagram, TikTok, YouTube, Reddit, Facebook, LinkedIn, Pinterest, and Threads).

How accurate is ScanTheFake?

ScanTheFake achieves up to 98.7% accuracy for image detection. However, AI detection is probabilistic — results should be interpreted as one data point among many. Accuracy varies depending on content quality, format, compression, and the sophistication of the AI model used to generate the content.

Can ScanTheFake detect ChatGPT, Claude, or Gemini text?

Yes. Our text scanner analyzes writing patterns including perplexity, burstiness, and stylistic markers to detect text generated by ChatGPT, Claude, Gemini, and other large language models. You can paste text directly or upload PDF, Word, or plain text documents for analysis.

What social media platforms does the link scanner support?

The link scanner supports 9+ platforms including Twitter/X, Instagram, TikTok, Facebook, LinkedIn, YouTube, Reddit, Pinterest, and Threads. It auto-detects the content type from the URL and performs the appropriate analysis (image, video, or text detection).

How to check if an image is AI generated?

To check if an image is AI generated, upload it to ScanTheFake's Image Scanner. Our AI analyzes pixel patterns, compression artifacts, metadata, and model-specific fingerprints to determine if the image was created by DALL-E, Midjourney, Stable Diffusion, Adobe Firefly, or other AI image generators. Results include a confidence score and the likely AI model used.

Can ScanTheFake detect AI-generated music from Suno and Udio?

Yes. Our Audio Scanner uses advanced acoustic pattern analysis to detect AI-generated music from Suno, Udio, Riffusion, and other AI music generators. Upload any audio file (MP3, WAV, OGG, FLAC, M4A) and get instant results showing the probability the music was AI-generated.

ScanTheFakeAI Detector

Pricing

Video Scanner

Upload a video to detect deepfakes and AI-generated footage

Drop your file here, or browse

MP4, MOV, WEBM — up to 100MB, max 3 minutes