How to extract text from a scanned PDF

There are two completely different things people call a PDF. One contains text: real characters, with fonts and coordinates, that you can select with your cursor. The other contains a photograph of text: a flat grid of pixels that happens to look like writing. To your eyes they are identical. To software they have nothing in common.

Every ordinary PDF converter, including ours, reads the first kind. It walks the text objects the file already contains and writes them out. When you hand it the second kind, there are no text objects to walk, so there is nothing to extract. The tool is not broken and your file is not corrupt: the text you can see was never stored as text.

Open the PDF in any viewer and try to select a line of text with your mouse. If you get a normal text selection highlight that follows the words, you have a text PDF and a converter will work. If nothing highlights, or a translucent box covers the whole page at once, you have a scan.

A second check: press Ctrl+F (Cmd+F on a Mac) and search for a word you can clearly see on screen. A text PDF finds it. A scan reports no results, because there is no text to search.

OCR stands for optical character recognition. Instead of reading text the file already has, it looks at the shapes in the image and guesses which letters they are. Modern OCR is very good on clean, straight, high-resolution scans of ordinary printed text. It gets noticeably worse with handwriting, unusual fonts, low resolution, skewed pages, coffee stains, and dense tables.

This matters for your expectations: OCR output always needs proofreading. It is a very good first draft, not a faithful copy. Numbers are the usual trouble spot, because a misread digit looks perfectly plausible and will not be caught by a spellchecker.

1. 1
   Try the tool already on your computerOn Windows, OneNote can OCR a pasted image (right-click the image, "Copy Text from Picture"). On macOS, Preview and even Quick Look let you select text directly on many images thanks to built-in Live Text — try selecting the text before assuming you need anything else. On iPhone and Android, the built-in camera or Photos app will usually let you select text in a picture.
2. 2
   Use Google Docs, which OCRs for freeUpload the PDF to Google Drive, then right-click it and choose "Open with → Google Docs". Drive runs OCR during the conversion and gives you an editable document. This handles multi-page files and costs nothing. Do not use it for confidential material you are not willing to upload.
3. 3
   Run Tesseract locally for private documentsTesseract is the long-standing open-source OCR engine. It runs entirely on your own machine, so nothing is uploaded. It is a command-line tool, so it is the least friendly option here, but it is the right answer for sensitive files: install it, then run "tesseract input.png output" for an image, or pair it with a PDF renderer for multi-page files.
4. 4
   Check whether the sender has the originalThis is the step everyone skips and it is by far the best outcome. A scan is usually a printed copy of a file that started life as a Word document, an invoice from an accounting system, or an export from a database. One message asking for the original file saves you OCR errors entirely.

Mixed files are common: a report generated as text with a signed page photographed and appended. Run our converter first. It extracts everything from the real text pages and skips nothing that exists, and the page count in the result tells you how much came through. Then OCR just the image pages and stitch the two together. That way you only proofread the pages that actually needed guessing.

If the selection test showed real text, you are in the easy case and there is nothing more to think about. Our extractor runs in your browser, keeps the reading order and paragraph breaks, and never uploads the file.