PDF to Word

PDF to Word builds a DOCX from extractable PDF text. Optional Page N headings (Heading 2) separate pages. Detected column groups can become Word tables.

• The PDF has a selectable text layer.
• You need a starting DOCX to edit, not a screenshot of pages.

• Turn a text report into {base}.docx for editing in Word.
• Keep Page N headings so you can find the matching PDF page.
• Use Clean when you want continuous prose without those page labels.

For each page, getTextContent() items go through extractPageContentBlocks. Headings, paragraphs, and tables are written with the docx package. Packer.toBlob produces {base}.docx.

Input

One PDF, up to 50 MB.

Output

{base}.docx

Headings

includePageHeadings defaults to on unless you turn it off.

1. Upload a text PDF.
2. Pick Paged or Clean.
3. Click Convert to Word.
4. Open {base}.docx.
5. Search for the no-text placeholder and check tables against the PDF.

Fonts, exact coordinates, headers/footers, and vector art are not reproduced as in the PDF. Table detection is best-effort. Scanned pages without a text layer do not become editable prose here.

pdf.js and docx run in the browser. The PDF is not uploaded for Word conversion. Identified analytics may include byte counts, page count, and whether page headings were included, not the filename or extracted text.

• PDF to Excel builds XLSX sheets from the same text-layout helpers, with different sheet modes.
• PDF OCR uses Tesseract.js on rendered page images and downloads .txt or .md, not DOCX.

• Search for the italic no-text placeholder.
• Check table column splits; detection is best-effort from pdfTextLayout.
• Compare a complex page to the PDF.
• If the file is a scan, run PDF OCR for TXT/Markdown instead of expecting this DOCX to invent text.