PDF to Word Converter

How it works

1. Drop in a PDF that has real text in it, not a scan.
2. Optionally name the pages you want.
3. The text is extracted and paragraphs are rebuilt from the page layout.
4. Download a .docx you can open and edit in Word.

About this tool

Turn a PDF back into a Word document you can edit. Be clear about what that means: this recovers the text and its paragraph breaks, not the design. A Word file describes a document, saying this is a heading and this is a table. A PDF describes a page, putting glyphs at coordinates. Going forwards discards meaning and keeps appearance; coming back has to infer the meaning from the appearance. Paragraphs can be inferred well from the spacing. Fonts, images, columns and tables cannot, and anything claiming otherwise is doing heavy processing on its own servers with your document. If you need the words so you can edit them, this works. If you need the original look, render the pages as images instead.

Why is this harder than it sounds?

A Word file describes a document: this is a heading, this is a paragraph, this is a table with three columns. A PDF describes a page: put these glyphs at these coordinates. Going from Word to PDF discards meaning and keeps appearance; going back has to guess the meaning from the appearance. Paragraph breaks can be inferred reasonably well from vertical gaps, which is what this does. Reconstructing a table from the fact that some text happens to line up in columns is a much harder problem, and tools that claim to do it well are running substantial processing on their own servers, with your document.

What you get

• All the text, in reading order for ordinary single-column documents
• Paragraph breaks inferred from the spacing on the page
• Optional page breaks where each PDF page began
• A real .docx that opens and edits in Word, Pages, LibreOffice or Google Docs

What you do not

• Fonts, sizes, colours and emphasis
• Images and anything drawn rather than typed
• Tables as tables, rather than as lines of text
• Headers, footers and footnotes, which appear inline where they sat on the page
• Reliable order in multi-column layouts, which can interleave

My PDF came out empty

Then it is a scan: a photograph of a page, where the words are pixels rather than characters. There is genuinely nothing to extract. Recognising text in an image needs OCR, which is a different and much heavier job. If you cannot select the words with your cursor in a PDF reader, no extraction tool can find them either.