Is my PDF uploaded to a server?

No. pdf.js and the docx writer both run entirely in your browser. Your PDF never leaves your device — verifiable in the Network tab. This matters for confidential contracts, signed agreements, medical records, financial statements, and any other private documents where third-party uploads are not acceptable.

Does the Word file look exactly like my PDF?

No. The output is reflowable text — paragraphs, font sizes, and page breaks are preserved, but exact layout, columns, tables, images, headers, and footers are not reconstructed. PDF is a fixed-page visual format; Word is a flowable text format. High-fidelity PDF→Word conversion that preserves exact layout is only available through commercial engines (Adobe Acrobat, Aspose, Nutrient) or AGPL-licensed tools like PyMuPDF. ToolMint focuses on letting you edit the words; for visual fidelity, use Adobe Acrobat.

Why are tables, images, and columns missing?

pdf.js exposes the text and its coordinates on each page, not a structural model of tables or layouts. Reconstructing tables and multi-column layouts from raw coordinates requires complex heuristics that no client-side library currently does well. Images are stored as separate streams in the PDF and would each need to be extracted, scaled, and embedded into the .docx — a non-trivial pipeline. The lite text path is what the browser ecosystem can deliver today without AGPL dependencies.

Can I convert scanned PDFs?

Not directly — pdf.js only extracts text that is encoded in the PDF, not text inside scanned images. For scans, run ToolMint's OCR PDF tool first to add a real text layer to the document, then convert the OCR'd PDF to Word here. The OCR tool runs Tesseract.js entirely in your browser and supports many languages.

Are headings and styles preserved?

Font size from the source PDF is mapped onto each Word run, so visually larger text in the PDF stays visually larger in the .docx. Headings are not tagged as Heading 1, Heading 2 in Word's style system because PDFs do not carry that semantic information — they only have font size and weight. After conversion, you can manually apply Word styles to large lines if you need a proper outline tree (Word's Navigation pane, Table of Contents).

What is the file size limit?

PDF uploads are capped at 50 MB. The bottleneck is browser memory during pdf.js page extraction — a 50 MB PDF with hundreds of pages of dense text uses substantial memory in the docx serialisation step. For very large PDFs, split into chapters first using ToolMint's Split tool, convert each chapter, then merge the resulting .docx files in Word.

PDF to Word

Extract PDF text into editable Word documents

Processed locally — your files never leave your browser

How It Works

Upload

Select or drag & drop your PDF. The file stays on your device — it is read into your browser's memory only. Both text-based PDFs (created by Word, LaTeX, or any digital tool) and scanned PDFs are accepted, but only text-based PDFs produce useful Word output. Scanned PDFs need OCR first.

Convert

pdf.js extracts the text content of every page along with each line's position and font size. The docx package (MIT) then assembles those lines into Word paragraphs, preserving font sizes and inserting page breaks between PDF pages. The result is a real .docx file you can edit in Word, Google Docs, or LibreOffice.

Download

Save the generated .docx directly to your device. Open it in Word, Google Docs, LibreOffice Writer, or any compatible editor. Page breaks from the source PDF are kept so the editable structure mirrors the original document.

Frequently Asked Questions

Related Tools

Word to PDF

Convert Word documents to PDF

Excel to PDF

Convert Excel spreadsheets to PDF

PowerPoint to PDF

Convert PowerPoint presentations to PDF

CSV to JSON

Convert CSV data to JSON