Processed entirely on your device — nothing is uploaded
How to use the pdf to word
- 1Drop your PDF into the box above.
- 2Choose what to save: a Word document, Markdown, HTML or plain text.
- 3Decide whether wrapped lines should be joined into paragraphs — leave this on for prose, off for anything already laid out line by line.
- 4Press Convert and download the file. The summary tells you how many words and headings came back.
Why PDF to Word is genuinely difficult
It is worth understanding what this conversion actually involves, because it explains both what to expect and why every tool that does it produces slightly different results. A PDF does not contain paragraphs. It contains instructions to draw strings of glyphs at particular coordinates on a page. There is no record of where one paragraph ends and the next begins, which line is a heading, or which lines form a bulleted list — that information was thrown away when the PDF was made.
So converting to Word is not decoding, it is reconstruction. This tool reads every text fragment with its position and type size, groups fragments into lines by their baseline, and then infers structure from geometry: a vertical gap larger than a line and a half ends a paragraph; a line that stops well short of the right margin followed by one starting with a capital ends a paragraph; a line set noticeably larger than the body text becomes a heading; a line beginning with a bullet character or a number becomes a list item; a hyphen at the end of a line before a lowercase letter is treated as hyphenation and joined back into one word.
These heuristics are good, not perfect, and they are good in proportion to how ordinary the document is. A report, a letter, an article or a book chapter converts very well. A magazine spread with three columns and a pull quote, a bank statement, or a form built from a grid of boxes will need editing afterwards — the words will all be there, but the reading order across columns is a genuinely hard problem that no converter solves reliably.
The two switches that matter most
"Join wrapped lines into paragraphs" is the important one. In a PDF each visual line is separate, so without this the .docx contains one paragraph per line — technically accurate and painful to edit, because every attempt to reflow the text fights the hard breaks. With it on, lines are joined into real paragraphs that reflow properly in Word. Turn it off deliberately for things that are meant to be line-by-line: addresses, poetry, code listings, tables of figures, subtitles.
"Turn larger text into headings" compares each line's type size against the document's median size — the median rather than the average, because a few large headings drag an average upwards and make everything else look small by comparison. Lines more than about 18% larger become headings, with the level chosen by how much larger they are. The result is a Word document with a real outline, so the navigation pane works and a table of contents can be generated.
Rebuilding lists is the third switch, and it recognises bullets, numbers, letters and roman numerals, including continuation lines that are indented under their item. The output is a proper Word list, not a paragraph that happens to begin with a dash — which means renumbering works when you edit it.
Scanned PDFs contain no text at all
If your PDF came from a scanner, a photocopier or a phone camera, its pages are photographs. There is no text layer to read, and no converter — this one, or any paid one — can extract words from it without optical character recognition first. The usual symptom of trying is a Word file that is completely empty.
This tool checks for exactly that and says so: it counts pages with no text layer and, if any are found, tells you how many and points you at the OCR tool on this site, which runs Tesseract in your browser in thirteen languages and produces text you can then convert. That is an extra step, but it is an honest one — a converter that returns a blank document and no explanation wastes far more of your time.
A mixed document — a report with a scanned annexe — is handled the same way: the text pages convert normally and the warning tells you which pages came back empty.
Why the output is a real .docx
Some converters produce a file with a .docx extension that is really an HTML document in disguise, or a Word file where every line is a text box positioned absolutely. Both open in Word and both are miserable to edit — the first loses styling as soon as you touch it, the second cannot be reflowed at all.
What this writes is a genuine Office Open XML package: a content-type map, package relationships, a document body, a style sheet defining the heading, quote and code styles, and a numbering definition so that lists behave as lists. It opens unchanged in Word, LibreOffice, Pages and Google Docs, and everything in it is editable in the normal way — you can restyle a heading and every heading at that level follows.
The alternative outputs exist because a Word file is not always what you want. Markdown is the right choice if the text is heading for a wiki, a static site or a repository. HTML is right if it is going into a CMS. Plain text is right when you only want the words. All four come from the same reconstructed model, so the structure the tool recovered is preserved in whichever you pick.
Frequently asked questions
Is my PDF uploaded to a server?
No. The PDF is read in this browser tab using pdf.js, and the Word document is assembled in memory and handed straight back to you as a download. Nothing is transmitted, which is the point — the documents people convert are usually contracts, statements and reports.
Will the Word file look identical to the PDF?
No, and no converter can promise that. You get the text with its structure — headings, paragraphs, lists — as an editable document. Exact page layout, columns and precise positioning are not reproduced, because that information is not stored in a PDF in a recoverable form.
My converted document is empty. Why?
Your PDF is almost certainly a scan, so its pages are images with no text layer. Run it through the OCR tool on this site first to produce real text, then convert that. The tool detects this case and tells you rather than silently returning a blank file.
Why is every line its own paragraph?
The "Join wrapped lines into paragraphs" option is switched off. Turn it on and consecutive lines will be joined back into paragraphs that reflow properly in Word.
Are tables converted?
Table text is recovered as text, but the grid is not rebuilt, because a PDF stores table borders as ordinary drawn lines with no link to the words inside them. If your PDF is mostly tabular data, converting to Excel via a CSV export from the source system is usually a better route.
Can it open a password-protected PDF?
Not one with an open password — that content is encrypted and unreadable without the key. Remove the password first with the Unlock PDF tool if you know it. Files with only a permissions password will usually convert without trouble.