Getting text back out of a PDF
PDF is a final format. It was designed to fix a document's appearance so it survives being sent anywhere, and it does that by recording where every character sits on the page. What it does not record is the structure underneath — which characters formed a sentence, where one paragraph ended, which line was a heading.
That is why editing a PDF directly is awkward, and why converting one back to Word is a reconstruction rather than a translation. This tool reads the text layer with pdf.js, groups the glyph runs into lines, works out where paragraphs and headings were, and writes the result as a real .docx you can open and edit.
What it rebuilds
Paragraphs. Lines that belong to the same paragraph are merged back together, using the gap between baselines, whether the previous line reached the right margin, and how the line ends. Words hyphenated across a line break are rejoined.
Headings.The most common text size in the document is taken as the body size, and anything meaningfully larger becomes a Word heading — mapped to the built-in Heading 1, 2 and 3 styles, so Word's navigation pane and automatic tables of contents work.
Lists. Bullets and numbers at the start of a line are recognised and turned into real Word list paragraphs, with each list getting its own numbering so a second list does not carry on counting from the first.
Page size.The .docx inherits the PDF's page dimensions, so an A4 document opens as A4 rather than defaulting to US Letter.
Repeated headers and footers
A running head or a page number appears on every page of the PDF, and in a re-flowed document it lands in the middle of your text every few hundred words. The converter looks at the top and bottom of every page, finds the lines that repeat across most of them — comparing with digits masked out, so "Page 4 of 30" matches "Page 5 of 30" — and removes them, telling you how many it took out.
It only does this on documents of three pages or more, where the pattern is meaningful, and you can switch it off if the repeated line is content you want to keep.
What does not come across
Tables become text rather than Word tables. Images, charts and drawings are not carried over. Exact fonts, colors and positions are replaced by ordinary Word styles. Multi-column pages are read in the order the PDF stores them, which for an academic paper often means reading across both columns rather than down one.
None of that is an oversight — those are the parts a PDF genuinely does not describe well enough to recover. The output is a working draft to edit, which is what people converting a PDF to Word almost always want.
Not OCR
This reads the text that is already in the file. If your PDF is a scan or a photograph of a page, there is no text in it — a person reading it on screen is doing character recognition with their eyes, and the computer has no equivalent unless something has run OCR and written a text layer back in. The tool detects that case and says so rather than handing you an empty document.
Common uses
- Editing a report whose original Word file is long gone
- Updating a contract or template you only have as a PDF
- Reusing text from a paper without retyping it
- Getting a CV back into an editable form to tailor it
- Pulling a long document into a format you can track changes in
Privacy
Everything runs on your device. The PDF is parsed in the tab and the Word file is assembled in memory — nothing is uploaded, stored or logged. Given what tends to be locked inside a PDF, that is the point.