Convert PDF to Word
Rebuild your PDF as an editable Word document with real paragraphs, headings, and tables. 100% private — never uploaded.
100% Private: All processing happens in your browser. Your files never leave your device.
This tool rebuilds a text-based PDF as an editable Word document, entirely on your device. It reads each page with pdf.js, rejoins lines that were wrapped mid-sentence back into real paragraphs, works out which blocks of text are headings, and reconstructs tabular regions as genuine Word tables with aligned columns — then hands the result to the docx library to produce an actual .docx file, up to 100 MB, one file at a time. Because the output is real text rather than a picture, it reflows: change the font size or margins in Word and the paragraphs re-wrap normally, and headings use Word's own heading styles so the navigation pane and an automatic table of contents both work.
The trade-off for that editability is fidelity to the original page. Your source PDF's fonts, colours and exact positioning are not carried over — the document is rebuilt from its text content, not redrawn pixel-for-pixel — and images are not extracted. A PDF that produces no readable pages has nothing for this tool to build a document from — most often because the pages are a scan or a photograph with no text layer, but the same empty result can also mean the file is damaged or protected, and the on-screen message says so rather than guessing a single cause. Use this converter when you need to actually edit or reuse a PDF's wording — a report, a notice, a résumé — not when you need a visual copy of a complex, image-heavy layout.
How it works
1. Drop or browse your PDF
Upload one text-based PDF at a time, up to 100 MB. There's no separate scanned-vs-text check at this step — any PDF is accepted here; the text-layer check happens once you press Convert.
2. Choose whether to keep bold and italic
“Keep bold and italic text” is on by default. It reads each page's embedded font data to work out which words were emphasised in the original. Turn it off to skip that lookup and speed up conversion of a long document — you'll get correct paragraphs and headings either way, just without bold/italic marks.
3. Press Convert to Word
You'll watch it move through Loading PDF…, Parsing document structure…, Reading page N of M…, then Rebuilding paragraphs, headings and tables…, Building Word document… and Saving document… — all in your browser tab, page by page.
4. Download the .docx
Once it reports Conversion Complete with a summary of pages, paragraphs, headings and tables converted, press Download Word (.docx). The file keeps your PDF's original name with the extension swapped to .docx.
When you'd use it
Editing an official notice or circular
A government or college notice distributed as PDF often needs to be adapted into a new letter, form annexure or internal memo. Converting it first gives you real paragraphs to cut, edit and restructure instead of retyping the whole thing.
Reworking a résumé you only have as PDF
If a placement cell, recruiter or an earlier version of yourself only left you a PDF résumé, convert it back to Word to update roles, dates and skills for a new application without rebuilding the document from scratch.
Adding comments or track changes to a report
Reviewers who work in Word rather than a PDF annotator need an actual .docx. Converting a text-heavy report first lets them insert comments and tracked edits directly in the paragraph flow.
Reading a long PDF comfortably on a phone
Because paragraphs and headings are rebuilt as real text rather than a fixed page image, the converted document reflows to fit a phone screen — useful for a long PDF report that's unreadable zoomed out on a small screen.
Text PDFs vs scanned PDFs
This tool can only work with a PDF that has an actual text layer — one where the words were typed or typeset, not photographed. If no page produces any readable content, conversion is rejected with a message stating that no pages could be produced and that the file may be empty, damaged, or protected — it doesn't assert which, since a scan, an encrypted file and a corrupted one all look identical from the inside: units went in, no text came out. If your pages are actually scans or photos, running an OCR tool first provides the text layer needed here; if the PDF is protected instead, unlock it first, then convert the unlocked copy here.
Some PDFs are mixed — mostly real text with one or two scanned pages mixed in (a signed cover page inserted into a typed report, for example). In that case conversion still succeeds, but you'll see an amber notice saying some pages had no readable text; those pages are simply skipped rather than the whole conversion failing, so you get a usable .docx for the rest of the document.
What becomes editable, and what doesn't
Paragraphs, headings and tables come through as real, editable Word content: wrapped lines are rejoined so text reflows normally, heading-styled text becomes actual Word Heading styles (so the navigation pane and an auto-generated table of contents both work), and regions of the page that are genuinely tabular are rebuilt as Word tables with their columns aligned.
What doesn't come through: your original fonts, colours, images and the PDF's exact visual positioning. The tool reconstructs the document from its text content rather than reproducing the page as designed, so a document with a complex visual layout — brochures, infographics, multi-column magazine-style pages — is a poor fit; a text-heavy letter, report or CV converts far more cleanly.
Fixing a reflowed layout after conversion
Because the tool rebuilds reading order rather than copying page coordinates, a multi-column PDF can come out with the columns concatenated in an unexpected order. If that happens, open the .docx in Word, turn on Show Formatting Marks (Ctrl+Shift+8) to see the actual paragraph breaks, and manually reorder or merge paragraphs into the sequence you want — it's usually a few minutes of cut-and-paste rather than a full retype.
For tables, check column widths and any merged header cells after conversion: cell text and row order are preserved, but exact widths are not, since the tool aligns columns rather than measuring the source PDF's layout in points. Re-sizing columns in Word afterwards is quick because the data itself is already correct.
Frequently asked questions
Why was my PDF rejected with no pages produced?
No page in the file produced any readable content. The on-screen message indicates the file may be empty, damaged, or protected rather than asserting a single diagnosis, because scans, corruption, and encryption all result in zero readable text. If your pages are scans, process them with OCR first; if the PDF is protected, unlock it before converting.
It converted, but warned that some pages had no readable text — what happened?
A few pages in your PDF had no text layer (often a scanned signature page mixed into an otherwise typed document) while the rest did. Those specific pages were skipped; the rest converted normally into your .docx.
Will bold and italic text be preserved?
Yes, if you leave “Keep bold and italic text” checked — it reads each page's embedded font data to detect emphasis. Turning it off skips that step for a faster conversion of long documents, at the cost of losing bold/italic marks.
Are tables converted into real Word tables?
Regions of the page that are genuinely tabular are rebuilt as actual Word tables with aligned columns and preserved cell text and row order. Column widths are not copied exactly from the PDF, so you may want to resize columns afterwards.
Why doesn't the converted document look exactly like my PDF?
The tool rebuilds your PDF's text as editable paragraphs, headings and tables rather than redrawing the page image, so fonts, colours, images and exact positioning are not carried over. That trade-off is what makes the result genuinely editable rather than a picture.
Is there a file size limit?
Yes — 100 MB per PDF, and one file at a time. Very long documents simply take longer, since each page is read and processed in turn in your browser tab.