How to make a scanned PDF searchable

First, check whether the PDF really needs OCR

Open the file and try to select a word with the mouse, then press Ctrl+F and search for a word you can see on the page. If the word highlights, the PDF already has a text layer and needs no OCR. If nothing can be selected or found, the pages are images.

PDF Information gives a second opinion: a document made of scans typically lists no fonts and reports embedded images, while a text PDF lists its fonts. Some files mix both, with text pages and scanned pages in the same document, which is why the OCR tool has an option to skip pages that already contain text.

What a searchable PDF is

A searchable PDF keeps the original page image and places an invisible layer of recognized text behind or over it, aligned to the words. You still see the scan, with its stamps, signatures and handwriting, but search, selection and copy now work because the text layer exists.

The text is only as good as the recognition. A misread character is invisible on screen but will make that word unfindable, which is why the quality of the source matters so much.

Prepare the scan for better recognition

Most OCR errors are caused by the image, not the software. Small corrections before you start save a lot of time afterwards.

• Scan at 300 DPI. Below about 200 DPI, small characters lose the detail that OCR relies on.
• Keep pages straight. A skewed page produces crooked lines that are harder to read.
• Use high contrast: dark text on a light background, no shadows from a book spine, no fingers in the frame.
• Avoid heavy compression or very low-resolution photos of documents.
• If you are still capturing the pages, Scan to PDF builds the file from your photos in the right order.

Run OCR PDF in your browser

OCR PDF works with Tesseract.js, so the pages are rendered, recognized and rebuilt on your device. Your documents are not uploaded. The first time you use it, the OCR engine and the language data are downloaded once from a public CDN, a few MB in total.

• Open OCR PDF and add your scanned PDF (JPG and PNG images are accepted too).
• Choose the language of the document. You can select several if a page mixes languages, for example English and French.
• Pick the resolution, 200 or 300 DPI. Use 300 for small print, 200 to finish faster.
• Keep the option to skip pages that already have text if the file is mixed.
• Choose the output: a searchable PDF, a plain-text file, or both, then start.

Language and script limits

The available languages are English, French, Spanish, Arabic, German, Italian, Portuguese, Dutch, Turkish and Russian. Choosing the wrong language is the fastest way to get poor output, because the engine matches shapes against that alphabet.

The invisible text layer inside the searchable PDF supports Latin-script languages. For Arabic and Russian, the tool gives you the recognized text as a .txt download instead, and the searchable PDF layer is not available for those scripts. Handwriting is not reliably recognized in any language.

Speed, size and what to expect

OCR is slow compared with most PDF tasks. Expect seconds per page on a laptop and longer on a phone, so a 200-page book can take a while. Keep the tab open while it runs, and try a few pages first to check the settings before committing to the whole document.

Because the page image is kept, the result is not a smaller file. If it turns out too large to send, Compress PDF can reduce it afterwards, and it is worth testing that search still works after compression.

Check the result

Open the new PDF, press Ctrl+F and search for three words: one from the first page, one from the middle and one from the end. Try a word with an accent or a number too, since those are the characters OCR confuses most, such as 0 and O or 1 and l.

If you only need the words rather than the layout, PDF to Text extracts the text layer into a .txt file. For a document that must be edited, run OCR first and then PDF to Word, because Word conversion has nothing to read on a scanned page.