Convert from PDF
What does PDF to Word do?
PDF to Word converts one PDF in the browser into a DOCX file. Extractable text becomes editable, positioned runs. Pictures and other non-text artwork stay as a background image on each page. Scanned pages have no editable text until you run OCR. Fonts and complex layouts can shift, so review the Word file before you rely on it.
How to use PDF to Word
- 1Choose one PDF
Open PDF to Word and select a single PDF from this device. Password-protected files must be unlocked first. The interface accepts one PDF per task.
- 2Check the preview
Confirm the pages you intend to convert. There is no extra layout control; conversion uses a fixed-layout mapping from each PDF page to a Word section.
- 3Convert locally
Start the conversion. Extractable text is placed in positioned frames. Non-text graphics are retained as page artwork behind that text.
- 4Download the DOCX
Open the file in Word or another editor. Check names, spacing, images and page breaks. Manual correction may be required on dense or uncommon layouts.
- 5Use OCR when text is missing
If a page is a scan or the glyphs were drawn as images, run OCR PDF first, then convert again so the searchable layer can be extracted.
Supported files and limits
- Output format: DOCX
- Input: PDF
- Output options: automatic fixed-layout conversion
- one PDF per task in the interface; manual correction may be required
- Files stay on this device and are removed when you close the page.
The output is a fixed-layout DOCX, not a native flowing Word document. Exact font metrics are not preserved, scanned pages need OCR first, and one PDF is accepted per task.
What the output keeps
- extractable text
- approximate text position and styling
- non-text page artwork
Text that PDF.js can extract is kept as editable Word text with approximate position, size and style. Non-text page artwork is kept as a background image so charts and photos remain visible.
What may change or be lost
- exact font metrics
- complex layout relationships
- editable vector graphics
- text from scans without OCR
Vector drawings are not rebuilt as Word shapes. Columns, tables and wrapping can shift. Text that exists only as pixels, outlines or an unsupported encoding is omitted until OCR or another source is used.
When not to use this tool
Do not use PDF to Word when you need a reflowing manuscript, pixel-identical print output, editable native drawings, or a slide deck. It is the wrong tool for scans that have not been through OCR, and it will not reconstruct spreadsheet formulas from a printed table.
If processing fails
If the PDF is encrypted, unlock it first. If the browser cannot parse the file, repair it or export a readable copy. If the DOCX opens with pictures but no text, the page was likely a scan or outline art; run OCR PDF and convert again. Very large files can fail on low-memory devices — close other tabs and retry with fewer pages.
Privacy boundary
The selected PDF is read in this browser for the active task. PDF Smart Kit does not send the document contents to an application server. Previews stay in memory until you close or refresh the page. The downloaded DOCX is stored only where your browser saves downloads.
Testing evidence
Last tested: · Test browser: Desktop Chrome via Playwright Chromium · Product version: 1.0.0
Sample: Digital PDF with extractable English headings · 2 pages
What was checked
- The download is a valid DOCX that opens in an editor.
- Extractable source text is present as editable Word text, not only as a picture.
- Non-text page artwork is retained as images.
- Output page count matches the source PDF.
Known failures and limits
- Scanned or outline-only pages have no editable text until OCR PDF is used first.
- The DOCX is a fixed-layout approximation; wrapping, fonts and native drawings can differ from the PDF.