Don't trust everything you see. AI-powered misinformation detection that analyzes images, videos, and social media content for signs of manipulation.
CheckKaro uses Claude Vision AI to detect:
- AI-generated content — Identify deepfakes and synthetic media
- Manipulated screenshots — Spot edited social media posts from Twitter, Instagram, and more
- Edited photos — Detect signs of image manipulation and tampering
- Fake video content — Frame-by-frame analysis of video for inconsistencies
- Upload an image, video, or paste a URL
- AI analyzes the content for signs of manipulation
- Get a trust score with detailed findings
- Image Analysis — Drag & drop any image for instant authenticity check
- Video Analysis — Frame extraction and multi-frame AI analysis via ffmpeg.wasm
- URL Analysis — Paste a Twitter/X or Instagram link for direct analysis
- Social Media Detection — Platform-specific analysis for screenshot authenticity
- Reverse Image Search — Find original sources via SerpAPI
- Trust Score — Clear 0-100 score with detailed breakdown of findings
- Framework: Next.js 14 (App Router)
- Language: TypeScript
- Styling: Tailwind CSS (black & white theme)
- Icons: Heroicons
- AI: Anthropic Claude API (Vision)
- Video: ffmpeg.wasm (client-side frame extraction)
- Search: SerpAPI (reverse image search)
# Clone
git clone https://github.com/imsaif/checkkaro.git
cd checkkaro
# Install dependencies
npm install
# Run locally
npm run devOpen http://localhost:3000.
MIT