What this tool does
Upload a PDF and get back an editable Word (.docx) document. The converter reads the text in your PDF, works out where the paragraphs, headings, tables and images belong, and rebuilds them as real Word content you can edit — not a picture of each page. Everything happens in your browser, so the file never has to be uploaded anywhere.
A worked example
Take a small invoice whose PDF contains a header row (Item, Qty, Price) and two product rows. The converter notices that the words line up in the same columns on every row, and rebuilds them as a genuine Word table of 3 rows by 3 columns, with the bold first row marked as the header. A PDF has no concept of a "table" — only text placed at coordinates — so this is a reconstruction based on alignment, and it works when the alignment is regular.
What converts well
Documents made from ordinary text convert best: letters, reports, articles, contracts, invoices and forms with simple layouts. Paragraphs, wrapped lines, headings, bold and italic text, text in other alphabets (accented Latin, Cyrillic, Greek and more), simple tables, embedded images, clickable web and email links, and page breaks are all carried across.
How it decides what is a heading
The converter compares each line with the size of the body text in the whole document. A short line that is clearly larger becomes a heading, with bigger text becoming a higher-level heading. A short line that is only slightly larger counts as a heading only if it is entirely bold. Long text is never turned into a heading, however large it is, and when the signals are unclear the line stays an ordinary paragraph.
Limitations
This is a reconstruction, not an exact copy, and some things do not survive:
- Tables are rebuilt only when they form a regular grid. Merged cells, irregular layouts and nested tables come out as plain text rather than a table, and a table whose cells contain long sentences may be read as columns of text instead.
- Multi-column layouts are reconstructed when the columns are clean and separated by a clear gap. Complex layouts — text wrapping around images, columns of very short lines, or unusual page designs — may be read in the wrong order.
- Bold and italic are detected from the font's own name. Some embedded fonts do not carry that information, so their styling can be missed. Font families, exact spacing, indentation and alignment are not preserved.
- Links are kept when the linked words can be identified exactly, which is the usual case for documents made in a word processor. Some PDFs store a whole line of text as a single piece, and then a link inside that line is left as plain text — you will see a note when this happens.
- Lists come across as paragraphs that begin with the bullet or number. They are not converted to Word's own list formatting.
- Images are carried across, but images larger than about 16 megapixels are left out, and some image types may not be supported. You will see a note when something is left out.
- Scanned PDFs — pages that are only pictures of text — contain no text to convert. Run them through an OCR tool first.
- Password-protected PDFs cannot be read; remove the password first.
Very large files are limited to 25MB and 300 pages, and a conversion that takes too long is stopped rather than left running.
Frequently asked questions
Will my Word document look exactly like the PDF? No. The aim is an editable document with the right content in the right order, not a pixel-for-pixel copy. Expect to tidy fonts and spacing, especially for designed documents.
Why did my table turn into plain text? Most likely its rows or columns were not aligned consistently, or it contains merged cells. Only regular grids are rebuilt as tables.
Why is there an image missing? Very large images are left out, and a note appears after conversion when that happens.
Can it convert a scanned document? Not on its own. A scan has no real text, so use an OCR tool to create a searchable PDF first.
Privacy
Your PDF is converted locally in your browser. This converter does not upload your file or send its contents anywhere, and nothing is stored after you close the page.