OcxlyDev · Free Tool

PDF to Word

Convert a text-based PDF into an editable .docx — keeping headings, paragraphs, and bold/italic. Everything runs in your browser, so your file never leaves your device.

Built for text-format PDFs — files whose text you can select and copy. Scanned pages, photos, and image-heavy or complex tables won't convert well, because this tool reads text, not pictures.

Drop a PDF here

or click to browse · processed locally, never uploaded

PDF

document.pdf

—

What to expect. This converter is built for text-format PDFs — documents whose text you can select and copy. Plain text, headings, and bold/italic come through reliably. Images and tables are best-effort and frequently don't survive: a PDF stores no real "table," "paragraph," or "figure" — only positioned glyphs and lines — so complex layouts, image-heavy tables, and scanned (image-only) PDFs convert poorly or carry no text at all. Password-protected or encrypted PDFs can't be opened either — remove the password first, then convert. This tool reads selectable text; it does not perform OCR. It works well on text documents; for richly designed or scanned files, expect to fix things by hand.

Please use this tool lawfully, and only on files you have the right to convert or edit. OCXLY accepts no responsibility for any unlawful, infringing, or unauthorized use of converted or edited documents.

§ 01

How the conversion works

A PDF and a Word document are fundamentally different things. A .docx is a flow of structured content — paragraphs, headings, tables, runs of bold or italic text — that reflows as you edit. A PDF is closer to a printout: a set of glyphs, lines, and images placed at fixed coordinates on a page, with no inherent notion of "this is a heading" or "these cells form a table." Converting one to the other means reconstructing that lost structure by reading the geometry.

Reading the page

This tool uses Mozilla's PDF.js to pull every text fragment out of your file along with its position, size, and font. It then rebuilds structure with a few heuristics: fragments sharing a baseline become a line; lines with consistent spacing become a paragraph; a line whose font is noticeably larger than the body becomes a heading; and a font name containing "Bold" or "Italic" sets the matching style on that run. The reconstructed content is written as genuine Word XML (the same OOXML format Microsoft Word saves), so the result is a real, editable document — not an image of one.

Tables

Tables are the hardest part, because a PDF table is just text with gaps and (sometimes) lines. The converter looks for runs of consecutive lines that split into the same columns — detected from wide spaces and aligned positions — and rebuilds them as native Word tables with borders. Clean, grid-style tables usually come through well. Borderless tables, merged cells, and nested tables are where heuristics struggle, and you may need to fix a few cells by hand.

Images

Embedded raster images are extracted at their original resolution, re-encoded as PNG, and placed inline in reading order at roughly the size they appear in the PDF. Vector graphics, background watermarks, and image masks may not transfer. The tool reads images already inside the PDF; it cannot recover text from a scanned document, which is a picture of a page rather than real text — that would require optical character recognition (OCR), which this tool does not do.

Privacy

Your PDF is opened, parsed, and converted entirely inside your browser tab using JavaScript. No part of the file is uploaded, logged, or sent to any server — the conversion works with no network connection at all, and nothing is retained once you close the page.