OCR PDF Free – Make Scanned PDF Searchable (No Upload)

How to make a scanned PDF searchable

A scanned PDF is a set of pictures of pages: you can see the words, but you cannot search, select or copy them. OCR adds the missing text.

1. Add your scanned PDF, or photos of the pages. On a phone you can take the photos directly.
2. Choose the languages in the document. The page language is suggested, but your document may differ.
3. Pick the pages. Pages that already contain text are detected, so only scanned pages are read unless you choose otherwise.
4. Optionally improve the scan for OCR: grayscale, more contrast, straightening or noise reduction. The preview shows exactly what OCR will read.
5. Run OCR, check the recognised text page by page and correct mistakes.
6. Download the searchable PDF, the text as a .txt file, or copy the text.

What is OCR?

OCR (optical character recognition) finds letters in an image and turns them into real text. This tool uses Tesseract, an open-source OCR engine, compiled to run inside your browser.

In the searchable PDF, the recognised words are stored as an invisible layer exactly over the words on the scan. The page looks the same as before, but your PDF viewer can now search it, and selecting a word highlights it on the scan. For example, an invoice scanned as images becomes searchable for its invoice number, and you can copy the total instead of typing it.

How to improve OCR results from phone photos

OCR reads printed text best when the letters are sharp, dark and evenly lit. A few habits make a large difference:

• Use daylight or bright, even light, and avoid the shadow of your phone or hand across the page.
• Hold the phone parallel to the page so lines are straight and letters are not stretched at one end.
• Fill the frame with the page and tap to focus; blurry photos lose small print first.
• Put the page on a dark, plain surface so its edges and lines are easy to find.
• If a photo is slightly tilted, turn on straightening. If it is faded, try more contrast; for grainy scans, try noise reduction.

Why can the PDF become larger after OCR?

Pages of a PDF you add are kept as they are, so the searchable PDF is usually only slightly larger: the text layer and a subset of the Inter font add a small amount per document.

Photos are a different case. JPEG and PNG photos are embedded unchanged (their location and camera metadata removed), but WebP and HEIC photos have to be converted to JPEG, which is often larger than the efficient original. If you need a small file, set a maximum size: scanned pages are then saved again as compressed images while the text stays searchable. The tool will not shrink text below a readable resolution just to reach a size.

Can OCR recognise handwriting?

No. The OCR engine is trained for printed and typed text. Handwritten notes are usually not recognised or come out garbled, so handwriting is not supported. Printed notes, forms, letters and receipts work well when the photo is sharp.

Is OCR private?

Yes. Your PDFs, images, the recognised text and your corrections stay in your browser. Nothing about your document is sent to a server.

The first time you use a language, its OCR language data (about 0.6–2 MB) downloads from this website, together with the OCR engine and the PDF reader. These downloads contain no information about your document, and the language data is then kept in your browser for future use.

Which languages are supported?

You can choose up to three of: English, German, Italian, Albanian, Greek, French, Spanish, Turkish, Serbian (Cyrillic or Latin), Croatian and Portuguese. Choose every language that appears: a German letter with English product names reads best with German first and English second.

Accented letters are part of the language data, so German umlauts and ß, Albanian ë and ç, and Italian accents such as è and à are recognised and stay searchable in the PDF.

What if the PDF already contains text?

Many PDFs already have real text, for example documents saved from a word processor or scans that were OCR'd before. Before reading, the tool checks every page: pages with selectable text, pages that are only an image, and mixed or uncertain pages are counted separately.

When every page already has text, OCR is unnecessary: the tool suggests extracting the existing text instead. For mixed documents it reads only the scanned pages and leaves the others untouched, so their original text is never duplicated. A page that only has a page number or a short header still counts as scanned.

Free online OCR without uploading documents

This is an online tool in the sense that you open it on a website, but the work happens offline from our point of view: your document is processed by your own browser and never uploaded to a server. There is no signup, no daily quota and no paid plan.

Processing time depends on your device. A few pages take seconds; long documents can take several minutes, and very large files may exceed a phone's memory.