Convert a PDF into a document you can edit in Word

Read this before you run it: what comes back is an HTML document with a .doc extension. Word, LibreOffice Writer, Pages and Google Docs all open one, and a single ‘Save as’ turns it into a real .docx — but this tool does not write .docx itself. What it gives you is the words in editable paragraphs with the headings marked up, which is what you want for rewriting a document and no use at all for reproducing one.

It does not write .docx files. What you get is an HTML document named .doc — Word, LibreOffice Writer and Pages all open it, and one ‘Save as’ turns it into a real .docx.

Convert to an editable document

No account. Your upload is deleted after 2 hours, the result after 24 hours.

How it works

  1. Choose the pages and the options. All of them or a range. Leave heading detection on unless the document is typeset in a single size throughout, in which case it has nothing to work with.
  2. Open the .doc it gives you. Double-click the download. Word will say the file is in a different format from the one its extension claims — that is this file being honest about being HTML — and open it when you say to.
  3. Save it as .docx from Word. File, Save as, Word Document. From that point it is an ordinary Word document with styles, track changes and everything else Word does.

What the options do

The form above lists the options this tool is running today, straight from the server. This is what they are for.

Pages to convert
Which pages to read, as 1-3,7 or ‘all’. Worth setting: converting the appendices of a 200-page report when all you wanted was the summary produces a document nobody can navigate, and every extra page is another chance for the paragraph reconstruction to guess wrong.
Find the headings
Type size is the one piece of styling a PDF states plainly, so it is what the headings are found from. The size most of the characters are set in becomes body text — weighted by how much text sits at each size, so a title page cannot define ‘normal’ — and lines set about 1.6 times that or larger become Heading 1, 1.3 times Heading 2 and 1.12 times Heading 3. A line longer than 200 characters is never treated as a heading however large it is set, because a pull quote in the navigation pane is worse than no heading at all. Turn it off for a document set in one size, where every guess would be noise.
Keep the page breaks
Starts each PDF page on a new page in the Word document. Keep it on when the pages mean something — a form, a set of certificates, a signed schedule. Turn it off for continuous prose, where the PDF’s page breaks fall in the middle of sentences and Word should be left to reflow the text.

What this tool does not do

  • It does not write .docx files. What you get is an HTML document named .doc — Word, LibreOffice Writer and Pages all open it, and one ‘Save as’ turns it into a real .docx.
  • Images do not come across, and neither do tables, columns, exact fonts, colours, headers, footers or anything else that is page furniture. Treat the result as text to edit, not as a copy of the page.
  • It cannot read a scan. A scanned page has no text layer, so the job stops with a message; the OCR tool reports which pages are affected.
  • Paragraphs are reconstructed from line spacing and from lines that stop short of the right margin. Unusual typesetting will have paragraphs joined or split in the wrong places.
  • Lists come back as ordinary paragraphs. The bullets and numbers are text that was on the page, not Word list formatting.

What happens to your file

The file is uploaded because PDF to Word runs on our server rather than in the browser tab. It is kept for at most 2 hours and the result you download for 24 hours, then a scheduled job deletes both — a job has to be able to read the file while it runs, and you have to be able to retry a failed conversion without re-uploading. Neither needs longer.

Nothing about the contents is logged, and no one reads your documents. The retention page lists the same numbers the code enforces, and the privacy policy says who can reach the storage they sit in.

Questions

Is this a real .docx converter?
No, and it says so on every run. It writes an HTML document with a .doc extension, which Word, LibreOffice Writer and Pages open directly and can save as .docx in one step. Writing OOXML packages — styles, numbering, section properties, relationships — is a much larger piece of work and is not part of this tool.
Word says the file is not in the format its extension says.
That warning is correct, and harmless: it is HTML in a file named .doc. Choose to open it anyway. Word is telling you it checked, not that the file is damaged.
Where have my tables and images gone?
They are not in the output — this converts text and heading structure only. For a table, PDF to Excel finds the columns and writes CSV; for the pictures, extract images pulls the embedded ones out and PDF to JPG renders whole pages.
The document came back as one long block with no headings.
Then everything on those pages is set at about the same size, which is all the tool has to go on: a PDF does not record ‘this is a heading’. The result says so in a note, and you will have to apply Word’s heading styles yourself.
Can it convert a scanned document?
No. A scan is a picture of words, so there is nothing to convert and the job stops rather than returning an empty document. The OCR tool will tell you which pages have no text layer.
Why do my paragraphs break in odd places?
Paragraphs are rebuilt from two clues a typesetter leaves: the gap between lines, and a line that stops short of the right margin. Justified text, deep indents and multi-column layouts confuse both. Turning off ‘keep the page breaks’ helps when it is the page boundaries that are splitting them.