Secure Client-Side PDF OCR

Extract text from scanned PDFs and photos of documents with OCR in 29 languages, including Chinese and Thai. Runs in your browser — nothing is uploaded.

Loading the tool…

Extract text from scanned PDFs with OCR

A scanned document is a picture of text: you can’t search it, select it or copy from it. OCR — optical character recognition — reads the letters in the picture and turns them back into text you can copy, edit and search. This tool reads every page of a scanned PDF, or photos and screenshots of documents, in 29 languages including Chinese, Japanese, Korean, Thai and Arabic.

How to OCR a PDF

  1. Upload a scanned PDF, or one or more images.
  2. Choose the language the document is written in.
  3. Click Start OCR. Each page is drawn at about 200 dpi and read in turn; you can stop at any time and keep the pages already read.
  4. Copy the text, or download it as a .txt file.

What you get

The text of every page, in order. When there is more than one page or image, each one starts with a line such as “--- Page 3 ---”, so you can find your place, and a summary tells you how many pages were read by OCR and how many already contained real text. The result is plain text: columns, tables, fonts and pictures are not kept. It is ideal for quoting a passage, searching an old contract, pasting into a spreadsheet or feeding a translator — rather than for rebuilding the document’s layout.

Supported languages

English, Spanish, French, German, Italian, Portuguese, Dutch, Polish, Czech, Romanian, Hungarian, Swedish, Danish, Norwegian, Finnish, Greek, Russian, Ukrainian, Turkish, Arabic, Hebrew, Hindi, Thai, Vietnamese, Indonesian, Japanese, Korean, Chinese (Simplified), Chinese (Traditional).

Tips for accurate results

  • Picking the right language matters more than anything else. For a document that mixes languages, choose the one most of the text is in.
  • Sharp, straight, evenly lit scans read best; blurry phone photos and handwriting read poorly. If you are scanning, 300 dpi in black and white or grayscale is plenty.
  • Pages that already contain real text are read directly — faster, and exact. Untick the option to OCR every page.
  • The first run downloads the recognition engine and the language data, so it takes a little longer.
  • Check names, numbers and amounts before you rely on them — OCR can confuse look-alike characters such as 0 and O, or 1 and l.
  • For a single image with clean-up controls, try Image to Text; to turn a text-based PDF into an editable document, PDF to Word; and to pull the text out of a PDF that isn’t scanned, PDF to Text.

Questions

Is there a page limit?

No. Long documents simply take longer, because every page is read in turn on your own device — and you can stop at any point.

Does it make a searchable PDF?

No — it gives you the text itself, ready to copy or save as a .txt file.

Recognition runs in your browser, so your documents are never uploaded — safe for contracts, IDs and medical records. Step-by-step guides: get text from a scanned PDF and OCR PDFs in other languages.

Guides

All how-to guides →

All PDF tools →