Map Extract
Free tool

Convert a PDF to an editable Word document

The agreement a client sent as a PDF needs one clause changing, or last year's proposal survives only as a PDF. Rather than retyping it, rebuild it as a Word file with proper headings and paragraphs you can edit.

Runs in your browser. Only if you turn on the optional engine (such as OCR) does it download open-source code from a public code CDN; your files stay on your device.

Example: reworking a service agreement

A web designer has a one-page service agreement as a PDF: a bold title, two bold numbered headings and short paragraphs whose lines wrap. We converted a page built like that with the Editable document preset.

Larger bold lines became headings and wrapped lines were joined back into whole paragraphs, so the DOCX edits like a document typed in Word, not like a page of separate lines.

Input
agreement.pdf, 1 page
Convert as: Editable document
Result
Pages 1, Headings found 3, Pictures 0, Scanned pages 0
Heading 1: Service agreement
Heading: 1. Scope of work
Paragraph: The supplier will design and build a five-page website for the client, including a contact form and a gallery of recent projects.
Heading: 2. Payment
Paragraph: Half of the fee is due on signature and half on launch.

Converting a PDF to Word

  1. Choose the PDF

    Conversion starts at once and a progress line counts the pages.

  2. Pick how to convert

    Editable document keeps pictures and flows text continuously; Page by page adds a page break after each PDF page; Text only drops pictures; Scanned PDF (OCR) reads image-only pages.

  3. Adjust and convert again

    Tick or untick Keep pictures, page breaks and OCR, then press Convert again with these options.

  4. Review the preview

    Figures show pages, headings found, pictures and scanned pages, with a preview of the rebuilt document underneath.

  5. Download

    Download Word (DOCX), or save the same content as Markdown or Plain text.

What converts well

Usually goodComes out as plain paragraphs
Headings detected from size and boldMulti-column pages
Paragraphs, with wrapped lines rejoinedTables
Bulleted and numbered lists, bold and italicText boxes and sidebars
Pictures, placed near where they appearedExact fonts and spacing

Scanned pages

Pages with almost no text are counted as scanned. Without OCR they produce no text. With OCR ticked, each such page is drawn and read with the open-source Tesseract engine in English, and its paragraphs are added to the document. Expect to proofread OCR text, especially names and figures.

Password-protected PDFs need unlocking with their password first, using Unlock PDF.

Questions people ask

How do I convert a PDF to Word for free?

Choose the PDF on this page, keep the Editable document preset and press Download Word (DOCX). The text is rebuilt into real headings and paragraphs in your browser, so there is no upload, account or watermark. Open the file in Word, LibreOffice or Google Docs to edit it.

Will the Word file look exactly like the PDF?

No, and that is deliberate. The layout is rebuilt so the text can be edited naturally, rather than copied as fixed boxes. Headings, paragraphs, lists and pictures carry over, but columns, tables and text boxes become ordinary paragraphs, and fonts and spacing follow Word's defaults.

Can I convert a scanned PDF to Word?

Yes, with OCR. Choose the Scanned PDF (OCR) preset or tick the OCR option. Each page without a text layer is read by a recognition engine and its paragraphs are added to the document. Results depend on scan quality, so check names, numbers and dates carefully.

Can I get the text without any formatting?

Yes. After converting, press Plain text for a .txt file with one paragraph per block, or Markdown if you want headings and lists marked up for a website, a wiki or a notes app. The Text only preset also leaves pictures out of the Word file.

Where your data goes

The PDF is read and rebuilt as a document in your browser and is never uploaded. If you tick OCR, the open-source Tesseract engine and its English data (about 5 MB) are downloaded once from a public CDN; your pages are still read on your own computer.

Checked against the tool on

Top