Drop a scanned PDF or photo here
or browse files — PDF, JPG, or PNG
First run downloads the language data (a few MB, then cached). Clear, high-contrast scans give the best results. OCR runs locally — your document still never leaves the device.
To extract text from a scanned PDF or photo for free: drop the file below, choose the language (English, Tamil, Sinhala, or mixed), and click Read text. Recognition runs locally with Tesseract — one of the few online OCR tools that handles Tamil and Sinhala without uploading your document anywhere.
or browse files — PDF, JPG, or PNG
First run downloads the language data (a few MB, then cached). Clear, high-contrast scans give the best results. OCR runs locally — your document still never leaves the device.
Most free OCR sites support only Latin scripts, and the ones that handle Tamil or Sinhala require uploading your document to a server. PaperPress runs Tesseract — the open-source engine originally developed at HP and Google — directly in your browser via WebAssembly, with trained models for both languages. That makes it suitable for digitising old letters, government forms, certificates, and textbook pages without privacy concerns.
English, Tamil (தமிழ்), Sinhala (සිංහල), plus English+Tamil and English+Sinhala mixed modes.
The language model (a few MB) downloads on first use and is then cached by your browser. The recognition itself always runs locally.
Use a clear, well-lit, high-contrast scan, keep the page straight, and pick the correct language mode.