{# Als einziges Symbol OHNE fill-Attribut: Der Stern zeigt seinen Zustand ueber die Flaeche (.sts-star[aria-pressed="true"] svg { fill: currentColor }). Ein fill am symbol waere ein Praesentationsattribut und schluege die geerbte CSS-Angabe - der gemerkte Stern bliebe dann eine Kontur. #} {# Bedienungssymbole des Notizblocks. Quelle: Lucide (ISC-Lizenz), wie in der Design-Uebergabe "Alltagswerkzeuge v2". Geometrisch statt handgezeichnet: sie stehen in Leisten neben Text, nicht als Kategoriezeichen. #}
Small Things Sorted

Convert PDF to Word

Text, headings, lists and tables back into an editable document.

Stays on your device – The conversion happens in your browser. Your files are not uploaded.

Drop your files here

or

You can also paste with Ctrl+V.

What you get is the content in editable form - text, headings, lists and tables. It is not a picture of the page: margins, column widths and the original fonts are set afresh by Word.

What you type is processed in your browser and never sent to us.

How to do it

  1. Pick the PDF

    Drop it in or choose it. Several at once is fine.

  2. Convert

    The file is read and written in your browser.

  3. Open in Word

    Headings, lists and tables are editable.

What this is for

A PDF is made to be read, not changed. Anyone who wants to fix one line faces a document built to prevent exactly that - and the source file usually belongs to somebody else.

What comes out is the content in editable form, not a picture of the page. A PDF stores no paragraphs, no headings and no lists, only single glyphs at coordinates. This tool reads the structure back out: the reading order, heading levels from the font sizes, lists, tables from the repeating column positions, and the bold and italic runs. The result is a real .docx carrying the Heading 1 to Heading 6 styles, so Word can build a table of contents from it.

Margins, column widths and the original fonts are set afresh by Word. Anyone who needs a pixel-perfect likeness does not need a Word file - they need the PDF.

There is nothing to take out of a scan: it holds no text, only a picture of each page. The tool spots that and says so instead of handing over an empty file.

Common questions

No. Both reading and writing happen in your browser. For contracts, references and invoices that is the difference that matters.

No, and that is deliberate. What carries over is content and structure - text, headings, lists, tables, bold and italic runs, and links. Margins, column widths and fonts are set afresh by Word. A pixel-perfect likeness would be a PDF again.

No. A scan holds no text, only a picture of each page. That needs character recognition, which is a different thing from reading a text layer. The tool spots the case and tells you.

If the columns are aligned, yes - a table column starts at the same position on every row, and that is what identifies it. A "table" that was merely spaced out arrives as text.

They are not carried over. Where an image belongs in the flow of text cannot be derived reliably from coordinates, and a misplaced image is worse than none. For the images there is the "PDF to images" tool.

No, they are removed. A line that sits in the page margin identically on every page is a running head - in a Word document it would land in the middle of the text.

Missing a tool? Tell us which one.