# Convert PDF to Word

Canonical HTML: [https://pdfpix.com/guides/convert-pdf-to-word/](https://pdfpix.com/guides/convert-pdf-to-word/)

A PDF stores finished pages, not a document, so no converter reads a paragraph out of one; it infers paragraphs from where the text was drawn. PDF to Word does that inference in this browser tab, then reopens the .docx it wrote and counts the text before offering it. This guide covers that route, Word's own Open command, and the scans neither can handle.

## Steps

1. **Choose the PDF.** Open PDF to Word and drop a file on the page, or pick one. It is read in this browser tab and never uploaded.
2. **Let it read the geometry.** Every drawn run comes back with its box, its size and the face it used, so the structure is inferred from measurements rather than guessed from pixels.
3. **Read the receipt.** The panel reports paragraphs, headings, tables, pictures, links and anything it refused, so you know what the document carries before you open it.
4. **Check and download.** The new .docx is reopened with a separate reader and its text compared with the text the PDF held. Below 98% there is no download.

## Why a PDF has no paragraphs to convert

A PDF records instructions: put these glyphs at this position in this font at this size. It does not record that two runs belong to one sentence, that a line is a heading, or that four runs are a table row. Those are inferences. Every converter that turns a PDF into Word makes them, and the difference between a good result and a mess is how honestly it handles the cases where the geometry does not decide. Changing the file format from PDF to Word is the easy half. Recovering the structure is the work.

[PDF to Word](https://pdfpix.com/pdf-to-word/) makes them here, in the tab, from the drawn boxes: runs that share a baseline become a line, lines with matching left edges and spacing become a paragraph, a larger face at the top of a block becomes a heading, and a column pattern repeated across consecutive lines becomes a table. Where the evidence stops it keeps the text and drops the claim, so a table whose columns do not line up stays as paragraphs instead of arriving as a grid with cells in the wrong places.

## What the browser converter writes

The .docx carries real Word objects, not a picture in a wrapper. Each line becomes a paragraph, a heading or a table cell, with the face, weight, slant, size, alignment and indent the PDF used. Text colour comes across as run colour. A picture is copied out of the PDF rather than redrawn, so a stored JPEG arrives byte for byte. A line that repeats in the same place on every page becomes a real header or footer part, with its page number as a field Word renumbers.

### Hyperlinks

A hyperlink is the one thing a PDF states outright: a rectangle on the page with a web address attached, held apart from the text it covers. The tool reads those rectangles and makes the runs under them clickable. A run is cut where the link changes between the pieces the PDF drew it in, and a rectangle that covers less than most of a piece names nothing, so a link over three words of a sentence stays over those three words instead of swallowing the sentence. A link can also point at a page of the same PDF, the way a contents page does. Word cannot hold that as an address: it needs a bookmark on the page being named. The tool works the destination back to a page and a point, puts a bookmark in front of the nearest paragraph, and points the line at it. A destination no paragraph came near keeps its text and loses its link. A web address on a running head is counted rather than written, because a header is a separate part with its own address list; an anchor needs none, so it survives there.

### Form fields

A form field holds its answer in the widget, not in the page content, so a tool that reads the page alone finds nothing where the answers are. This one reads the fields. A text field, a check box, a radio button and a dropdown become Word content controls, with the answer already in them and the dropdown list intact, so the reader can still fill the form in. An empty field gets a rule of underscores cut to its own width and a check box gets a box glyph. Neither is text the PDF held, so both come out of the count before it measures anything. A push button and a signature field hold no answer Word could carry, so they are counted and left behind.

### What does not come across

Drawn vector art. A picture the tool cannot copy faithfully is left behind and counted rather than re-encoded into something worse. Rotated text is kept upright at the end of the page it came from rather than woven into a paragraph it does not belong to. Every one of these is a number on the receipt, so a page that lost something says so.

## Nothing is trusted on the way out

The finished .docx is reopened with the same OOXML reader the [Word to PDF](https://pdfpix.com/word-to-pdf/) direction uses, a reader that shares no code with the writer, and its non-space characters are counted against the characters the PDF held. At least 98% has to come back or no download is offered at all. A converter that cannot show you that number is asking you to proofread its output yourself.

Over a 6,338-file corpus the median document kept 100% of its characters, and the files that fell under the floor were refused rather than handed over. That is the measurement behind the claim, and it is the reason a refusal is a feature here and not a failure.

## How to change a PDF to Word in Word itself

Word opens a PDF directly. File, Open, pick the .pdf, and Word warns that it will convert the document and that the result may differ from the original. Accept, and Word runs its own inference and shows you an editable document. It is a good route when Word is already open and the PDF is simple, and it is the only route that also runs Word's own equation recognition.

It has three costs. It uploads nothing, but it does send the file through Word's converter, which on a document of any length is slow. It rewrites the page into Word's own layout, so the result often looks less like the PDF than a geometry-driven conversion does. And it needs Word. The browser tool needs a tab, works on a Chromebook, and never moves the bytes anywhere.

## What to do when you convert PDF to Word and get nothing

An empty or nearly empty document almost always means the PDF has no text layer. A page photographed or scanned is an image, and there are no drawn runs to read; nothing in the file says the pixels spell anything. This tool refuses that file rather than handing back an empty .docx, because a guessed transcription needs optical character recognition and its own accuracy measure, and we do not ship one yet.

Two ways to tell in ten seconds. Open the PDF in any reader and try to select a line of text: if the cursor drags a blue box over the whole page instead of highlighting words, it is an image. Or search the document for a word you can see. No result means no text layer. For those files use an OCR tool first, then convert the searchable PDF it produces.

A password-protected file is the other common refusal. Unlock it first, with [Unlock PDF](https://pdfpix.com/unlock-pdf/) if you have the password, and convert the unlocked copy.

## Why the Word file does not look identical

It cannot, and a converter that promises it will is describing a picture. A PDF holds fixed pages; a .docx holds flowing content that Word lays out again using the fonts on the machine that opens it. Page size, page breaks, paragraph shape and table columns are kept, but Word rebreaks the lines, and a face the PDF embedded that your machine does not have is substituted.

That is the trade you are making on purpose. A file that looks identical is not editable, and a file that is editable is laid out again. If what you actually want is the page exactly as drawn, you want the PDF, and [inserting the PDF into Word](https://pdfpix.com/guides/insert-a-pdf-into-word/) as an object or a picture keeps it that way inside a Word document.

## Other routes, and when to use them

Google Docs opens a PDF from Drive with Open with, Google Docs, which runs Google's converter and includes OCR on scans, in exchange for uploading the file to Google. LibreOffice Writer opens a PDF through its Draw import, which preserves position at the cost of putting every line in its own frame, awkward to edit but accurate to look at. Adobe's own export is the most faithful of the paid options and needs an Acrobat subscription.

Each of those is a different point on the same trade between fidelity, editability and privacy. The browser tool takes the position that nothing leaves the tab and the result is checked before you see it. Neighbouring guides, including the other direction in [Convert Word to PDF](https://pdfpix.com/guides/convert-word-to-pdf/), are under [PDF, Word and Excel help](https://pdfpix.com/guides/office/).

## Questions and answers

### How do I convert a PDF to Word for free?

Open PDF to Word, drop the file on the page, and download the .docx it checks and offers. It runs in the browser tab with no account, no upload and no limit on how many files you convert.

### Can I turn a PDF into Word without losing the formatting?

Fonts, weight, slant, size, colour, alignment, indents, tables, lists, headers, footers, pictures and links are kept. The line breaks are not: Word lays the text out again, which is what makes the result editable.

### Why is my converted Word document empty?

The PDF has no text layer. A scan is an image of a page, so there are no drawn runs to read. The tool refuses that file rather than returning a blank document. Run it through OCR first.

### Does it work on a scanned PDF?

No. This version needs native text. Guessed text needs optical character recognition and an accuracy measure of its own, and offering it without one would break the 98% check the rest of the tool rests on.

### Do hyperlinks survive the conversion?

Yes. Link rectangles in the PDF become real Word hyperlinks, cut to the words they cover. A link that points at a page of the same PDF becomes an anchor into a bookmark the tool writes on that page. A web address on a running header is counted on the receipt instead of written.

### Is Word or the browser tool more accurate?

They infer differently. Word rewrites the page into its own layout and reads equations; the browser tool follows the drawn geometry more closely and proves the text survived. Try the free one first.

### Does converting upload my PDF anywhere?

No. The file is opened, converted, reopened and checked in this browser tab. The bytes never leave it, which is also why there is no file-size queue and no account.

### What does the 98% check actually prove?

That the .docx holds the text the PDF held. The tool reopens the file it just wrote with a separate reader, counts the non-space characters, and refuses to offer a download below the floor.

## Related pages

- [PDF to Word](https://pdfpix.com/pdf-to-word/): Turn a native-text PDF into an editable Word document. The file stays in your browser. No upload, no account needed.
- [Convert Word to PDF](https://pdfpix.com/guides/convert-word-to-pdf/): Save a Word document as a PDF, with its hyperlinks and heading bookmarks intact. A guide, with the tool that does the job named in it.
- [Insert a PDF into Word](https://pdfpix.com/guides/insert-a-pdf-into-word/): Put a PDF inside a Word document. A guide, with the tool that does the job named in it.
- [Unlock PDF](https://pdfpix.com/unlock-pdf/): Save a copy of a PDF you can already open, without its password. The file stays in your browser. No upload, no account needed.
- [PDF, Word and Excel help](https://pdfpix.com/guides/office/): Find the right guide for PDFs inside Excel, Word and other software. Every guide on the job, sorted by the problem you have.