# OCR a PDF — Make a Scan Searchable and Selectable

> Recognise the text in a scanned PDF and add it as an invisible, selectable layer over the original pages.

A scanned PDF is a picture of a page: you cannot search it, select from it or copy out of it. OCR reads the pictures and puts the words back. This runs Tesseract in your browser, then draws each recognised word transparently at the exact position it was found — so the document still looks precisely as it did, keeps its original quality and file size, and is now searchable in any reader. Nothing is uploaded, which is the point when the scan is a passport, a contract or a medical record.

**URL:** https://convertto.tech/t/ocr-pdf
**Category:** PDF Tools (https://convertto.tech/c/pdf-tools)
**Privacy:** Runs entirely in the browser; no upload
**Cost:** Free, no sign-up
**Last updated:** 2026-08-01

## Key facts

- **Method:** A transparent text layer over the original pages — nothing is re-rendered
- **Languages:** 16 languages plus bilingual combinations
- **Privacy:** Runs entirely in your browser — nothing is uploaded
- **Cost:** Free, unlimited, no sign-up

## How to use

1. Select your scanned pdf — the file stays on your device and is never uploaded.
2. Enter or paste your pages.
3. Choose the language.
4. Set the recognition detail.
5. Set the discard words below.
6. Choose the produce.
7. Enter or paste your page plan.
8. Press Run, then download the result when it is ready.

## FAQ

### Does OCR change how my document looks?

Not at all. The original page objects are kept exactly as they were and the recognised words are drawn at zero opacity on top. Tools that rebuild the PDF from images instead will visibly soften a scan and multiply its file size.

### Why is it slow?

Recognition is genuinely heavy work and it is happening on your own machine rather than a server farm. Expect a few seconds per page. Start with a page or two to confirm the language is right before running a long document.

### The text is wrong in places.

OCR accuracy depends almost entirely on the scan. A clean 300 DPI scan reads near-perfectly; a phone photo at an angle, a fax, or a page with a coffee ring will not. Raising the recognition detail helps with small print, and choosing the correct language helps more than anything else — English models guessing at German produce exactly the nonsense you would expect.

### My PDF is not a scan and OCR found nothing useful.

Then it already has a text layer and needs no OCR. The PDF Text Extractor will read it directly, accurately and instantly.

### Is my document uploaded?

No. The recognition model is downloaded to your browser and the pages are processed there. The traineddata file is cached, so later runs in the same language start immediately.

## Sources

- [Tesseract OCR](https://github.com/tesseract-ocr/tesseract) — Tesseract

## Related tools

- [Crop PDF](https://convertto.tech/t/crop-pdf): Trim the margins of every page, or just the pages you choose, without re-rendering.
- [PDF Text Extractor](https://convertto.tech/t/pdf-text-extractor): Pull all the text out of a PDF as plain text you can search, copy or reuse.
- [Scan to PDF](https://convertto.tech/t/scan-to-pdf): Use your camera to capture pages and turn them into a clean, multi-page PDF.
- [Duplicate PDF Pages](https://convertto.tech/t/duplicate-pdf-pages): Repeat chosen pages any number of times — for labels, tickets, forms or handouts.
- [Edit PDF](https://convertto.tech/t/edit-pdf): Add text, pictures, shapes, arrows, highlights and freehand drawing to any PDF page.
- [Excel to PDF](https://convertto.tech/t/excel-to-pdf): Convert an Excel workbook to a PDF table, one section per sheet, in your browser.
- [Extract PDF Pages](https://convertto.tech/t/extract-pdf-pages): Pull specific pages out of a PDF into a new document, in one file or one file per page.
- [Fill PDF Forms](https://convertto.tech/t/fill-pdf): Read a PDF's interactive form fields, fill them in, and optionally lock the answers.