Account

What new tool do you want?

Document Converters

PDF to Word

Convert a PDF into an editable .docx with its paragraphs and headings rebuilt. Scanned pages come through as images.

How to use

  1. 1Choose a PDF.
  2. 2Press Convert. Text becomes editable paragraphs; larger text becomes headings.
  3. 3Open the .docx in Word, Google Docs or LibreOffice. Complex layouts (columns, tables) may need tidying.

Frequently asked questions

Will scanned PDFs convert to editable text?

No — scanned pages have no underlying text, so they come through as images in the .docx. Use an OCR tool first if you need editable text from a scan.

Will the formatting look exactly the same?

Paragraphs and headings are rebuilt, but complex layouts like multi-column pages or heavy tables may need manual tidying.

How paragraphs and headings get rebuilt from raw positions

A PDF has no concept of "this is a paragraph" or "this is a heading" — it only has lines of text positioned on a page. This converter reassembles structure by watching two signals: a large vertical gap between lines (or a change in font size) starts a new paragraph, while lines that sit close together at the same font size get merged into one, with a trailing hyphen treated as a broken word rather than a real hyphen. Headings are detected by comparing each block's font size against the median body-text size on the page — text roughly 1.6× the body size becomes a Heading 1, 1.3× becomes Heading 2, and 1.12× becomes Heading 3. This heuristic works well on conventionally formatted reports and documents, but a PDF with unconventional font sizing (e.g., bold body text that's barely larger than normal) can get misclassified.

Scanned pages become embedded images, not editable text

When a page has no extractable text at all — a photographed page, a fax, a flattened scan — rather than silently dropping that page, it's rendered as an image and embedded directly into the .docx at that point in the document, keeping your page count and reading order intact. That image is not OCR'd, though: it looks like a page from a document, but none of its text is selectable or editable in Word. If you need that text editable, running an OCR tool on the PDF first (to generate a real text layer) and then converting the OCR output here will produce real paragraphs for that page.

Where this struggles: columns and tables

The paragraph-reconstruction logic works line by line down the page, which assumes a single reading column. A PDF laid out in two newspaper-style columns doesn't have that structure encoded anywhere explicit, so lines from the left and right columns can get interleaved in reading order, producing garbled paragraphs. Dense tables face a similar issue — cell text becomes a sequence of short text blocks rather than a Word table, since there's no table-reconstruction logic here (that's what the dedicated PDF-to-Excel tool is built for, for tabular data specifically).

Example: a 9-page single-column report with two scanned appendix pages

Picture a 9-page PDF report: pages 1-7 are normal single-column text with section headings noticeably larger than body text, and pages 8-9 are scanned signature pages with no text layer. Converting produces a .docx where pages 1-7 come through as real, editable paragraphs with the section headings correctly promoted to Word's Heading 1/2 styles (so they show up in a generated table of contents), while pages 8-9 appear as embedded images in their original position — readable and printable, but their signatures and any handwritten notes aren't selectable text. The note shown after conversion ("2 scanned page(s) included as images") flags exactly which pages those were.

Related tools

More: Document Converters