PDF to Text (OCR)
Upload a PDF, get the text back - entirely in your browser.
Click to add a PDF or drop it here
Working…
Upload a PDF, get the text back - entirely in your browser.
Click to add a PDF or drop it here
Pick a language, then add a PDF (click, or drag and drop). Each page is first rendered to an image in your browser, then read with OCR the same way a photo would be - this runs entirely in your browser and starts automatically. Copy the result or download it as a .txt file.
No. Everything runs in your browser using Tesseract OCR compiled to WebAssembly - the file is never sent to a server. That also means it still works if you lose your connection partway through, as long as the page is already loaded.
Each page is first rendered to an image in your browser (via pdf.js), then read the same way a photo would be. Multi-page PDFs are processed one page at a time and the text is combined, with a page marker between pages.
The OCR engine and the language's trained data download once per language, then are cached by your browser for next time. A PDF with many pages also naturally takes longer than a short one.
OCR accuracy depends on scan quality and resolution - a low-resolution scan is harder to read correctly than a crisp one. The extracted text is editable in the box above so you can fix anything before copying it.
Yes, up to 30 pages at a time, so a very large PDF can't hang the tab.
8 visits