SnapOCR - Screenshot & Text Extractor

v1.3.1 - Coming Soon preview for Pro features • Pro upgrade button is now clearly labeled "Coming soon" and disabled • Removed all payment redirects — the Pro section is preview only, no payment is being collected in this version • Removed unused Lemon Squeezy host permission for a cleaner permission set • All free OCR features continue to work exactly as before ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ v1.3.0 - Focused release for free users • All core OCR features remain fully functional and free • Cloud OCR engines (Gemini / Qwen / OpenRouter) — bring your own key • Pro checkout is being finalized — leave your email on GitHub to be notified when it launches • 100% on-device OCR by default, no data leaves your browser unless you opt in • Under the hood: improved stability and minor UI polish ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 🆕 v1.2.0 UPDATE — Cloud AI Vision & Supporter Pro • Optional cloud AI vision engines (Gemini 2.5 Flash, Qwen-VL, OpenRouter) — bring your own API key, nothing sent by default • 3 more languages: Traditional Chinese, Japanese, Korean (now 5 total) • Supporter Pro: 5 premium themes + supporter badge ($9.9 one-time) • Smarter post-processing, cleaner UI, bug fixes ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ SnapOCR — Screenshot any region, get the text in one click. Now with optional AI vision for the hardest cases. ✨ WHAT IT DOES Draw a box on any webpage. SnapOCR instantly extracts the text inside and copies it to your clipboard. No more retyping quotes from images, PDF viewers, video subtitles, error dialogs, or scanned documents. 🔒 PRIVACY-FIRST BY DEFAULT The default engine (Tesseract.js) runs 100% inside your browser — screenshots and text never leave your device. No account required, no upload, no tracking. 🚀 NEW: OPTIONAL AI VISION FOR HARD CASES When Tesseract struggles (handwriting, low-contrast text, complex layouts), switch to a cloud VLM engine with your own API key: • Google Gemini 2.5 Flash (best accuracy) • Alibaba Qwen-VL / DashScope (works in mainland China) • OpenRouter aggregator (free tier with Llama Vision & Nemotron) Cloud engines only activate when YOU pick them — never by default. Your API keys are stored locally. ⚡ CORE FEATURES • One-click region capture on any website (works on canvas, video frames, protected PDFs) • Local OCR + 3 optional cloud VLM engines • 5 languages: English, Simplified Chinese, Traditional Chinese, Japanese, Korean • Smart post-processing: mixed-script detection, common OCR error correction • History panel with thumbnails — re-copy anytime • Auto-copy to clipboard (optional) • Lightweight (~300 KB), no bloat 💎 SUPPORTER PRO (Optional, one-time $9.9) Love the extension? Unlock 5 beautiful themes (Sakura / Forest / Sunset / Midnight / Ruby) and a Supporter badge. Buy me a coffee — no subscription, no recurring fee. All OCR features remain free forever. 🎯 PERFECT FOR • Copying text from images, memes, screenshots • Extracting quotes from PDFs and slide decks • Grabbing text from video subtitles (including anime, YouTube) • Copying error messages from dialogs that block selection • Digitizing scanned documents and old photos • Students, researchers, translators, developers 🛡️ PERMISSIONS EXPLAINED • "activeTab" — draws the selection overlay only on the tab you click. We don't read your browsing history. • "storage" — saves your OCR history LOCALLY on your device. • "clipboardWrite" — used only when you press Copy (or auto-copy if you opt in). • Host access to Google / Alibaba / OpenRouter / Lemon Squeezy APIs — used ONLY when you enable a cloud engine or activate a Pro license. Zero traffic by default. 📮 FEEDBACK Found a bug? Want more languages? Email 820619304@qq.com — we read every message.