OCR PDF (Make a Searchable PDF)

A scanned contract or a photographed form is just a picture: you can’t search it, copy a sentence or let a screen reader read it. OCR (optical character recognition) reads the letters on each page and adds them as an invisible text layer, so the PDF looks exactly the same but becomes searchable and selectable. Recognition runs in your browser; the document is never uploaded.

How to use this tool

1. Choose the scanned PDF.
2. Pick the document’s language — Korean + English, Japanese + English and Chinese + English cover mixed text.
3. Use 200 DPI for normal documents or 300 DPI for small print, and limit the pages if you only need some.
4. Wait while each page is recognised, then download the searchable PDF and, if useful, the plain text.

Formula

Each page is rendered at the chosen DPI (pixels = points ÷ 72 × DPI) and recognised by the Tesseract engine, which returns a text-only PDF page in an invisible font. That page is scaled to the original page size and drawn on top of it.

Pages displayed rotated are handled so the text layer lines up with what you see. Pages that already contain at least 5 characters of real text are skipped by default.

Example

A 12-page scanned lease at 200 DPI: each A4 page becomes a 1,654 × 2,339 px image for recognition, which takes a few seconds per page on a laptop. The result keeps the scans unchanged and adds the text layer, so searching for the landlord’s name finds every page that mentions it.

Download size and accuracy

The first run downloads the recognition engine and the language data — roughly 10–20 MB depending on the language — from a public CDN; after that the browser caches it. Accuracy is best on straight, evenly lit scans at 200–300 DPI with printed text. Handwriting, stamps, tables with tiny fonts and photos taken at an angle produce more mistakes, so check important names and numbers.