Word to PDF Converter

Why a converted document usually looks wrong — and what fixes it

Almost every browser-based Word converter asks you to pick a font, a size and a margin. That question is the bug. A .docx already carries all three: the font is declared once in `docDefaults` inside styles.xml, the size in half-points on each run, the margins in twentieths of a point on the section. A converter that asks you instead is telling you it never read them — it's re-typesetting your document from scratch with its own defaults, which is why the output never matches what you had on screen.

This one reads the file. Fonts, point sizes, colours, bold and italic, alignment, indents, spacing and line height all come from the document itself. The controls on this page are overrides for when you actually want to change something, and they're off by default — because a document you didn't ask to restyle shouldn't come back restyled.

What survives, what doesn't, and why

Word stores far more than text, and a PDF can only show what the converter understood. This is the honest split rather than a marketing one.

• Kept: headings and their style, bold/italic/underline/strikethrough, per-run fonts and sizes, colours and highlights, superscript and subscript, alignment, left/right/first-line/hanging indents, space before and after, line spacing, bulleted and numbered lists with their nesting, tables with their authored column widths and cell shading, inline images, and the page size and margins from the section.
• Approximated: fonts you have and the PDF doesn't. A PDF embeds its own fonts, and this build uses the standard set — so Calibri and Aptos map to a sans face, Cambria and Georgia to a serif, Consolas to a monospace. The layout holds; the letterforms are close rather than identical.
• Lost: headers and footers, footnotes and endnotes, comments, tracked changes, text boxes, shapes, charts and SmartArt. If your document leans on any of those, this isn't the tool — Word's own Save as PDF is.

The character problem nobody warns you about

The standard PDF fonts cover Latin text and little else. Feed them Devanagari, Chinese, Arabic, Cyrillic or Greek and there is no glyph to draw — the characters don't warn you, they simply don't appear. The same applies to some punctuation that looks harmless: certain dashes, arrows and mathematical symbols are outside the encoding too.

The conversion tells you when this happens rather than shipping a page with holes in it. If your document is not in a Latin script, use Word or LibreOffice to export it — they embed the real font and the text survives. This is a limitation of what can be embedded in the browser, not a rendering preference.

Getting a clean result the first time

• Check the page size before you convert. A document authored on US Letter and converted expecting A4 reflows nothing — the section's own size is used — but if you need the other one, set it in Word first.
• Images are embedded at their original resolution, so a document full of phone photos produces a large PDF. Shrink them in Word (Compress Pictures) or run the output through Compress PDF afterwards.
• A table wider than the text column is clipped, not shrunk. If a table runs off the page in Word, it will run off the page here too — fix it at the source.
• Going the other way? PDF to Word rebuilds an editable .docx from a PDF, with the caveats that direction has.