Image & PDF to Text Converter — Free Offline OCR
Turn photos, screenshots and scanned PDFs into editable text in 50+ languages. OCR runs on your device, so files are never uploaded.
Help shape Image & PDF to Text Converter — Free Offline OCR
The bug you just noticed will fade in 3 minutes. The feature idea will vanish after your next tab switch. Crazy peps don't suffer silently.
Frequently Asked Questions
Drop, paste (Ctrl+V) or pick a JPG, PNG, WebP, screenshot or PDF. Text is extracted automatically and appears on the right, ready to edit, copy, or download as TXT, Markdown, JSON or CSV. Click any highlighted box on the image to jump to that line.
No. Detection and recognition run inside your browser with WebAssembly. The only network request is a one-time download of the open-source model files. Your images, PDFs and extracted text never leave your device, so it is safe for invoices, IDs and contracts.
Yes. Drop multi-page scanned PDFs and batches of images together; every page is queued and read in order. Copy all pages at once or export them as one TXT, Markdown, JSON or CSV file.
English and most Latin-script languages, Chinese, Japanese and Greek work out of the box. Pick Korean, Russian and other Cyrillic languages, Arabic, Persian, Urdu, Hindi, Marathi, Nepali, Thai, Tamil or Telugu from the Language menu.
Yes. The first run downloads the model once — about 20 MB (Fast), 45 MB (Balanced) or 153 MB (Max Accuracy) — and stores it on your device. After that it loads straight from disk and OCR keeps working with no internet connection.
Baidu's open-source PaddleOCR (PP-OCRv6), converted to ONNX and run with ONNX Runtime Web. Every model file is pinned and SHA-256 verified. For photos or tiny print, choose Max Accuracy and set Detection resolution to Ultra.