PDF Tools
PDF to Word
Extract Unicode text from every digital-PDF page into an editable DOCX.
Updated
One PDF up to 12 MiB and 50 pages. Every page must contain extractable text. No OCR, image preservation, table reconstruction, or layout guarantee.
What this conversion does
The converter asks PDF.js for positioned text from every page, orders the text from top to bottom and left to right, and groups nearby lines into paragraphs. It then writes those paragraphs to an editable DOCX, using soft line breaks within a grouped paragraph and a page break between source pages.
What the DOCX preserves
The output preserves the extractable text returned by PDF.js, the converter's paragraph grouping, line breaks inside those groups, and boundaries between PDF pages. It is designed for recovering editable text, not recreating the original page design.
Expected output
A two-page digital PDF becomes a DOCX with the ordered text from page 1, a page break, and then the ordered text from page 2. Nearby source lines may share one Word paragraph with line breaks; the converter does not promise the same columns, spacing, or pagination as the PDF.
Limitations and common failure cases
Not reconstructed
Images, fonts, exact coordinates, columns, and visual styling are not written to the DOCX. Table, header, and footer structures are not reconstructed; text from those regions may still appear as ordinary extracted paragraphs. Multi-column or heavily positioned pages can produce a different reading order.
Files that can fail
Image-only or blank pages without extractable text cause conversion to fail; encrypted, corrupt, mislabeled, oversized, or over-limit PDFs are also rejected. A scanned PDF may work if it already has an extractable text layer.
The selected PDF is read and converted in your browser; this conversion path contains no document-upload request. Normal site resources may still load, so browser processing is not a confidentiality guarantee. Review the DOCX before editing or sharing it.
Frequently asked questions
Does it work with scanned PDFs?
It may work only when every page already contains an extractable text layer. This tool does not perform OCR, so an image-only page fails.
Will Word look exactly like the PDF?
No. The output is a text-focused DOCX, not a visual reconstruction of the fixed PDF layout.
Are images or tables preserved?
Images are not written to the DOCX. Table structure is not reconstructed, though extracted text from a table may appear as ordinary paragraphs.
Why can paragraph order differ?
Reading order and paragraph grouping are estimated from text coordinates and spacing, which can be ambiguous in complex layouts.
What happens to the download?
A browser object URL provides the DOCX until it is replaced, reset, or the component closes. Your browser controls where the file is saved.