Turn photos, screenshots and scanned PDFs into editable, copyable text with optical character recognition (OCR) — entirely in your browser. Drop a PNG, JPG or WebP, or a scanned PDF (pick a page range), choose the languages in the document, and the engine reads the text on your device. Copy the result or download it as a clean .txt file. Nothing is uploaded to a server, so it is safe for receipts, IDs and confidential documents.
Straightforward VPS and app hosting with predictable pricing.
How to extract text from an image or scanned PDF
Drop in a photo, screenshot or scanned PDF — the file is read locally and never uploaded to a server.
Select the language (or languages) in the document — English, Hindi, Telugu, Tamil and ten more are supported, and mixed-language pages can be selected together.
For a PDF, optionally enter a page range (e.g. 1-3,5); each page is rendered as an image, so scans with no text layer work too. OCR is capped at 20 pages per run.
Press Extract text, watch the progress bar, then copy the result or download it as a .txt file.
How OCR here works
The tool runs the Tesseract OCR engine compiled to WebAssembly, downloaded on first use from the same CDN the site already trusts for its video tools. Small images are upscaled before recognition because OCR accuracy climbs with resolution, and oversized photos are scaled down to keep memory use in check. The result is plain text: words in reading order, without fonts, colours or layout.
Getting the best results
Sharp, straight-on, high-contrast images read best. Blur, shadows, heavy compression and skewed pages are the usual causes of misread characters — re-shooting a photo flat-on helps more than any setting. Screenshots of digital text come out near-perfect; fancy display fonts and handwriting are the hardest cases. When the output comes back empty, the tool says so plainly instead of pretending nothing was there.