Drag-select text to confirm whether it is a scan.
Use the editable mode of PDF to Word; the server recognises text and rebuilds the document.
Verify amounts, IDs and names by eye rather than trusting the output.
A text PDF stores character codes: selectable, searchable, readable by screen readers. A scanned PDF stores one photograph per page — to software it is just pixels. They can look exactly the same on screen, which is why a tool that works for someone else seems to do nothing for you.
Try to drag-select a sentence. If text highlights, it is a text PDF. If nothing highlights, or the cursor is a hand everywhere, it is a scan. File size is a second clue: scans usually run 200KB or more per page and grow with page count.
OCR recognises characters in the image and places a text layer underneath. Quality depends on scan resolution, language and layout. Handwriting, stamp overlays and skewed phone photos degrade sharply — always spot-check the important numbers afterwards.
The Word file is rebuilt from recognised text, so page images are not preserved — keep a backup of the PDF if the appearance matters.
Noticeably worse than printed text. For important documents, type the key fields manually.
No. PDF to Word runs on our server, free and without a usage cap.