PP-OCR Chinese OCR

PP-OCR lets you select any visible area on a webpage and extract text using PP-OCRv6 — Baidu's state-of-the-art OCR model, running entirely in your browser via WebAssembly. No uploads. No API keys. No account required. HOW TO USE Click the PP-OCR icon (or use the keyboard shortcut) Drag to select any area on the page Text is extracted instantly and shown in a toast Click to copy KEY FEATURES • Chinese OCR optimized — PP-OCRv6 tiny model, tuned for Chinese + English • 100% local inference — ONNX models run in-browser via WebAssembly • Works on any page — screenshots, articles, images, PDFs rendered in browser • No internet required after installation • Lightweight — tiny det (1.5MB) + rec (4.3MB) models bundled PRIVACY All processing happens locally. Screenshots are captured in-browser, processed by WASM, and discarded — never sent to any server. Privacy policy: https://gist.github.com/chatgptnexus/113facf4cd228ae0b89b0b559fab4c10