Last reviewed: 2026-10-25
If your PDF came from a scanner or a phone camera, each page is a photograph of a document, not text. Here is how a scan becomes an editable Word file, what you get back, and how to give the OCR the best chance.
Open it and try to select a single word with the mouse. If you can highlight just that word, the PDF has real text. If the whole page highlights as one block, or nothing highlights at all, the page is an image. Searching with Ctrl+F (Cmd+F on a Mac) is another quick test: a scanned page finds nothing.
OCR (optical character recognition) looks at the picture of the page, finds the lines of text and works out which letters they are. Our converter uses OCRmyPDF, which runs the open-source Tesseract engine, and reads pages in English and Greek. Tesseract can handle more than 100 languages in general, but only the language packs installed on our server are used (see the Greek guide for what that means for other scripts).
For a fully scanned document you get the recognised text as normal Word paragraphs: headings set in larger type in the scan are made bold, lines are joined into paragraphs, words split with a hyphen at the end of a line are joined again, and each page starts on a new page. You can edit it like any document you typed.
What you do not get is the look of the original. Tables come back as rows of text with tabs between cells, logos and photos are not included, and fonts are Word’s default. That is a deliberate choice: until 2026-10-25 the converter tried to keep the page picture and laid the text over it in floating boxes, and the result looked like two copies of the page printed on top of each other.
In our test, a clean English scan at 200 dpi came out word for word; the same page at 100 dpi had one wrong word in 304; a copy tilted by 1.5 degrees with specks added was still perfect. A scanned Greek form reached 93.6%, with mistakes in a large bold title and in letters with diaeresis. Real scans with shadows, stamps, handwriting or faint print will do worse than our test pages.
Handwriting is not reliably recognised. Forms where the layout matters as much as the words (a filled-in government form you need to resubmit, for example) are usually better retyped into a fresh template, using the OCR text to copy from. And if your file mixes normal pages with scanned pages, split out the scanned pages and convert them separately, because OCR is only triggered when the first pages have almost no text.
Yes. OCR runs automatically and free on this site, with no sign-up. Files are deleted as soon as you download the result.
For fully scanned documents the converter returns the recognised text as normal paragraphs. Tables come back as tab-separated rows and images are not included, because laying text over the page picture produced unreadable results.
Not reliably. The OCR engine is built for printed text.