Cross-platform AI content detector. Paste a tweet, drop a screenshot, share a TikTok — one tap and you'll know whether it was made by a human or a model.
Most "AI detectors" check one thing. AI Analyzer checks the screenshots, photos, videos and posts you actually see in your feed — with analysis powered by Claude, an AI model by Anthropic, plus specialist image forensics.
Claude reads the writing the way a forensic editor would — rhythm, filler phrases, missing lived detail — and gives a clear verdict.
Claude vision + Sightengine flag Midjourney, DALL-E, Stable Diffusion and Flux images. Text inside a screenshot is read and checked too — two verdicts in one tap.
Key frames from your clip are checked together for AI-video tells: morphing objects, changing text, warping backgrounds, impossible motion.
Take a screenshot or share from any app — AI Analyzer offers to check your latest screenshot or copied text the moment you open it.
Content is hashed and cached server-side for 30 days at most. We never sell, never train on your uploads, never ship a creepy analytics SDK.
Every result comes with a short plain-English explanation from Claude — not just a number. You see exactly why.
We combine a frontier AI model with a specialist image detector and tell you how confident the result is.
iOS via TestFlight and Android via internal track are rolling out shortly. Drop us a line and we'll let you in.