Skip to main content

OCR — recognize text in scans

Turn scanned PDFs and images into searchable PDFs or text — right in your browser.

Drop a file here or choose one

PDF or image (JPG, PNG, WEBP, TIFF)

Recognize text in scanned PDFs and images (OCR) online

A scanned document or a phone photo is an image: you can't search, copy or edit its text. OCR (optical character recognition) reads the text from the image. Uzbek in Latin and Cyrillic script, Russian and English are supported.

The result can be a “searchable PDF” — the original page stays the same and an invisible text layer is added on top, so search and copy work. Or you get just the text as a .txt file. Recognition runs with Tesseract's accurate (“best”) models right in your browser.

How do I get text from a scanned PDF?

  1. Choose a scanned PDF or image(s).
  2. Tick the document language (several are allowed).
  3. Pick the output: searchable PDF or text.
  4. Click “Recognize” and download the result.

Frequently asked questions

No. Only the language model is downloaded once; recognition runs in your browser and the file goes nowhere.

Uzbek (Latin and Cyrillic), Russian and English. For a mixed document, tick several languages.

Very high on clean, straight scans. Blurry, skewed or low-quality photos produce more errors — scan at 300 DPI or higher when possible.

On first use the selected language model (4–15 MB) is downloaded. Next time it is already stored in your browser.