1. Home
  2. PDF
  3. PDF to Word

PDF to Word

Rebuild the text and tables of a PDF as an editable Word (DOCX) file.

Runs in your browser · No upload
This rebuilds the text and tables - it is not a server-grade converter like Word or Acrobat. Fonts, positioning, colours and shapes may not carry over, so always check the result by eye.

How to use

  1. Drop a PDF onto the box or choose one. If the PDF is password-protected, enter its open password.
  2. Pick a method. The default is Rebuild as text and tables (editable). Choose Keep the look if you’d rather keep the original appearance.
  3. Choose all pages or a range such as 1-3, 5, 8-. You can convert up to 300 pages at once.
  4. Click Convert to DOCX, then Download DOCX. The file is named original-name.docx.

The two methods

Rebuild as text and tables Keep the look
Result Headings, paragraphs, lists and tables as real Word formatting One 150 dpi picture per page
Editing and search Yes No (they’re pictures)
Appearance May differ from the PDF Same as the PDF
Scanned PDF Can’t be converted (no text) Works (pages become pictures)
File size Small Hundreds of KB per page

What this tool can and can’t do

This isn’t a server-grade converter like Word or Acrobat. It reads where each piece of text sits in the PDF, makes its best guess at the structure, and rebuilds the document from that. The result can look different from the original, so always check it by eye.

  • Headings: text at least 1.6 times the body size (the most common size) becomes Heading 1, at least 1.3 times becomes Heading 2, and at least 1.15 times, or a bold line on its own, becomes Heading 3.
  • Paragraphs: a gap between lines more than 1.5 times the usual one starts a new paragraph. Words split by a hyphen at the end of a line (exam- + ple) are joined back together.
  • Lists: lines that start with •, -, 1., ① or 가. become list items. The marker stays as ordinary text (it isn’t Word’s automatic numbering).
  • Tables: a table is only created when three or more lines in a row share the same column start positions (within 1.5% of the page width) and there are at least three columns. Uneven paragraphs aren’t turned into tables. Tables whose column starts shift around, such as right-aligned number columns, or tables with merged cells may be missed or come out as paragraphs.
  • Two-column pages: if an empty vertical strip runs down the middle, the left column is read first, then the right.
  • Bold and italic: only recognized when the font name contains Bold or Italic. In PDFs that hide font names, this formatting may be lost.
  • Pictures: when turned on, pictures that can be extracted as JPEG or PNG are added at the end of each page. Their original position on the page, vector drawings and background images behind text are not carried over.
  • Headers, footers, footnotes, fonts and colors: these get no special handling, so they end up mixed into the body text or dropped.

After converting, the tool reads the finished file back with mammoth and shows you whether the character count matches. That check only tells you no text went missing, not that the layout matches.

Scanned PDFs and copy restrictions

  • A scanned or photographed PDF has no text layer, so it can’t be converted to text. You’ll see a notice. To get the words, first run it through PDF OCR to make a searchable PDF, then convert that. To carry over only the look, use Keep the look.
  • If the author has restricted copying, you’ll see a warning, and you have to tick a box confirming you’re allowed to convert the document before it goes ahead.

If you only need the words, use PDF to Text. For tables, use PDF to Excel. To save pages as images, use PDF to JPG / PNG, and to pull out the embedded pictures, use Extract PDF Images. To open the result, try the DOCX Viewer. If a password stops the PDF from opening, use Unlock PDF first.

Your file is processed only in your browser and is never sent to a server.

Can't find the tool you need?

Tell us what you want to do. We build the most requested tools first.