How to convert a PDF to Word (and what won't survive)

Almost everyone converting a PDF to Word wants the same thing: to change some words in a document they no longer have the original of. A contract with the wrong date. A CV they wrote three laptops ago. A form somebody sent as a PDF because that was easier than asking what software you own.

That is a solvable problem. What is not solvable — and what every converter quietly glosses over — is producing a Word file that looks identical to the PDF. Understanding why takes two minutes and saves you an afternoon of disappointment.

Why a PDF and a Word file disagree fundamentally

A Word document describes intent. It says "this is a Heading 1, this paragraph follows it, this list has four items", and Word decides where the lines break when you open it. Change the margin and everything reflows.

A PDF describes a finished result. It says "draw the letter T at 71.3 points from the left and 604.8 from the bottom, in 11pt Calibri". There are no paragraphs in a PDF. There are no headings. There is not even necessarily a space character between two words — the gap may just be where the next letter happens to start.

So converting means reading a page of positioned glyphs and inferring the document that would have produced it. Words and their order come back reliably. Which lines were headings can be worked out from their size. The exact layout cannot, because the information describing it was never stored.

Converting a PDF to Word

1. 1Open the PDF to Word tool. Nothing uploads — the conversion runs inside your browser, which matters if the document is a contract or a payslip.
2. 2Drop the PDF in. Multi-page files are fine; each page break in the original becomes a page break in the .docx.
3. 3Leave heading detection on unless the result comes out with too many headings, which happens on documents set in a single size throughout.
4. 4Run it and open the .docx in Word, Google Docs, LibreOffice or Pages. It is a real Word file, not a renamed PDF.
5. 5Fix the layout, not the words. Reapply your own styles rather than nudging each paragraph — it is faster and the result holds together.

What comes across, and what doesn't

• Carries across: all the text, in reading order; headings, detected by size; bullet and numbered lists as real Word lists; page breaks.
• Does not: multi-column layouts, which flatten into one column; precise positioning; images; text inside tables keeps its words but not the table.
• Cannot: a scanned PDF has no text at all — see below.

If the goal is editing the words, the losses are cosmetic and you were going to restyle it anyway. If the goal is an exact replica of the design, no browser tool and no website will give you that, and the honest answer is to rebuild it from whatever the PDF was made in.

If your PDF is a scan, this won't work

A scanned or photographed document is a picture of words. There is no text in the file to convert, so a converter has nothing to work with and hands you an empty document — which looks like a broken tool and is actually a broken input.

What you need first is OCR: software that looks at the picture, recognises the letter shapes and writes a text layer behind them. Once a PDF has that layer, converting it works normally. A quick test: try selecting a sentence in your PDF reader. If you cannot highlight the words, there are no words.

Common questions

Will the formatting be exactly the same?

No, and any tool claiming otherwise is overselling. Text, reading order, headings and lists come across. Exact positioning, multi-column layouts and images do not, because a PDF does not record the structure that would let a converter rebuild them.

Is converting a PDF to Word free?

Here, yes — no account, no watermark, no page limit. The conversion happens in your browser using your own computer's processing, so there is no server cost to pass on to you.

Do I need Microsoft Word to open the result?

No. The output is a standard .docx, which Google Docs, LibreOffice Writer, Apple Pages and Word Online all open. Google Docs is the quickest route if you have no desktop office software.

Is my document uploaded to a server?

Not by this tool. The PDF is read and the .docx written inside the browser tab, which is why it still works with your network switched off. Most converters do upload, and for anything confidential that distinction is the whole story.

Why did my converted document come out as one long column?

Because the original was laid out in columns. A PDF stores where each piece of text sits, not that two blocks were meant to be read side by side, so a converter reading top to bottom flattens them. Newsletters and academic papers are the usual culprits.