Convert PDF to Word
Turn any PDF into an editable Word document — images included, with an Exact look mode that mirrors the original pages, and built-in OCR for scans. Nothing is stored; your file is never kept.
- Never stored
- No queue, no waiting
- No signup, no watermark
How it works
Add your PDF
Drop a PDF onto the page. The tool checks it automatically and switches OCR on if it detects a scan.
Pick the mode
Editable text, the default, gives you a normal Word document, page for page with the PDF: paragraphs, headings, bold and italics, links and images, each page at its original size with its own margins and line spacing, so page 12 of the PDF is page 12 of the document — in a file about the size of the PDF. Exact look pins every line where it was over a picture of the page — it mirrors the PDF for printing, but is hard to edit and several times larger. For scans, OCR switches on by itself, works out the language from the first page, and straightens pages that were scanned askew.
Convert and download
Click Convert to Word and download a clean .docx with headings and paragraphs, ready to edit.
What converts cleanly, and what does not
A PDF records glyphs at coordinates. It has no paragraphs and, unless someone deliberately tagged it, no headings — so both are worked out rather than read. Words sharing a baseline become a line, a widening vertical gap starts a new paragraph, and a short line set at about 1.6 times the document's median body size becomes a Heading 1, 1.3 times a Heading 2. That last rule is worth knowing before you start editing: a document whose headings are bold but no larger than the body text arrives as body text, and Word's navigation pane will be empty.
Links survive, which surprises people. A link annotation over a piece of text becomes a real Word hyperlink, and a contents entry or cross-reference that jumped to page twelve becomes an internal link to a bookmark planted at the top of that page — so a table of contents still works after the conversion.
Two things are not read at all. The answers in a filled-in PDF form live in field objects rather than in the page, so in Editable text they are simply not there; Exact look shows them, because they are part of the picture of the page, but not as text you can edit. Comments and stamps are the same kind of object and fare the same way. And in Exact look the artwork is rendered once at 144 DPI with the real text placed over it, so a logo or a chart that was vector in the PDF is a picture in the .docx, and enlarging it shows that.
When you want something else
When what you want is the words rather than a Word document, pdftotext -layout file.pdf from Poppler is faster and more predictable, holds columns apart with spaces, and is the usual way to feed a PDF into anything else. And if the PDF was made from a Word file in the first place, ask whoever made it for the .docx: every conversion back is a reconstruction from coordinates, and the original is the only copy that is not.
Tables are the weak point of every PDF-to-anything converter, this one included. If a document is mostly tables and the numbers matter, Tabula and Camelot are free, are built for that one problem, and will beat a general converter at it.
Frequently asked questions
Is my PDF stored when I convert it?
No. Both conversion modes store nothing — your file is never kept. With OCR, the reader and the language pack come from this site the first time; your document itself is never sent anywhere.
Can it convert scanned PDFs and photos of documents?
Yes. Switch on OCR (the tool suggests it automatically when it detects a scan) and the text is read from the page images. The language is detected from the page — over 120 are supported, including English, Bengali, Hindi, Urdu, Arabic, Spanish and Chinese — and you can pick it yourself if the guess is wrong. Ruled forms come out as real Word tables in Editable text mode.
How accurate is the OCR?
On clean scans of printed text, accuracy is very high. Quality drops with blurry photos, handwriting or unusual fonts. Clear, straight, well-lit scans give the best results.
Will my formatting and images be preserved?
Yes. Exact look keeps each page at its original size and rebuilds it in two layers: every photo, logo, colour band and piece of vector artwork is reproduced exactly as the PDF draws it, and the real text sits on top at its original position, with the original colours, bold and headings. Editable text instead rebuilds clean paragraphs and headings with images in reading order, page for page with the PDF — the better choice when you want to edit heavily.
What Word format do I get?
A standard .docx file that opens in Microsoft Word, Google Docs, LibreOffice and every modern editor.
Is there a page or file limit?
No fixed limit. OCR takes a few seconds per page since your own device does the reading, so very long scans simply take longer — a progress indicator shows exactly where it is.
Good to know: Fonts are substituted with close equivalents, because PDFs embed their own fonts and Word cannot use them — so Exact look is near-identical rather than pixel-identical, and a line may run slightly wider or narrower than the original. Tables in text PDFs convert as positioned text rather than Word tables; in scans, ruled tables become Word tables but unruled ones stay as positioned text. OCR deliberately ignores photos, logos and decorative graphics so they never become garbage characters, leaves handwriting and stamps as pictures, and takes a few seconds per page.
Put this tool on your website
Free for any blog, class page or help article. Paste one snippet and your visitors can use it right on your page.