DOCX to PDF Converter — Word to PDF In Your Browser, No Upload

Browser-local document converter

Use DOCX to PDF

Shard Tools recreates the DOCX layout locally in your browser and aims to keep the PDF visually close to the original. Complex Word features may still render differently. The file is not uploaded, and this site has no endpoint that could receive one.

Word document (.docx)

Check the PDF before you send itPage size, margins, headings, bold, italic, underline, colour, alignment, indents, lists, tables and inline images are reproduced. Fonts are the honest limit: a .docx names fonts such as Calibri or Cambria without carrying them, and a web page cannot read the licensed originals, so families are mapped onto the standard PDF fonts. Their metrics differ slightly from Word's, which means a line can break in a different place. Complex Word features may still render differently, so open the PDF and look at it before sending it anywhere.

Checked against a document whose headings used a custom style based on Heading 1, which a converter reading style ids alone would have flattened into body text — the basedOn chain is followed, so they stay headings and keep the inherited bold and size. A table nested inside a cell was also tested: it is reported rather than drawn, and critically the outer table's remaining rows survive, which is where a naive regex split truncates the parent table.

Explore document tools tools

Find related browser-local tools for nearby tasks without starting another search.

Browse all document tools tools

About DOCX to PDF

Turning a Word document into a PDF is the most ordinary task in office software, and it is also the one where people most often hand a confidential file to a stranger. The documents that need converting are contracts, offer letters, medical forms, tax paperwork and CVs, and the usual free converter accepts them by upload, processes them on a server somewhere, and hands back a link. This tool does the work in your browser tab instead. The .docx is unzipped here, its XML is read here, and the PDF is drawn here with a PDF library running in the page. Nothing is transmitted, and this site operates no endpoint that could receive a document. What you get is a real reconstruction rather than a text dump: the page size and margins from the document's own section properties, the paragraph and character formatting resolved through Word's style inheritance, lists with their real numbering, tables with their column widths and borders, and inline images redrawn at the size the document asked for. It aims to look close. It does not claim to be identical, and the section below is specific about where the differences come from.

Using DOCX to PDF

The .docx is opened as the ZIP archive it actually is, and word/document.xml is read for the body, word/styles.xml for the style definitions, word/numbering.xml for list formats and the relationship file for images, headers and footers.
Formatting resolves the way Word resolves it: document defaults first, then the whole style chain through basedOn, then any direct formatting on the paragraph or run. That chain is why a custom style built on Heading 1 is still recognised as a heading rather than being flattened into body text.
Page size, margins and orientation come from the document's section properties in twentieths of a point, so an A4 document produces an A4 PDF and a US Letter document produces Letter, without either being assumed.
Text is measured with the exact font that will draw it and wrapped at word boundaries, run by run, so a sentence that switches from regular to bold mid-line is measured in both faces rather than in one.
Tables are measured before they are drawn. A row that will not fit in the space left moves to the next page whole instead of being torn across the break, and column widths come from the document's own table grid, scaled down only if the grid is wider than the page.
Anything recognised but not reproducible is listed as a warning before you convert, so you can decide whether the conversion is good enough for what you are about to do with it.

Where docx to pdf helps

A report with headings, tables and a logo

The heading hierarchy, table borders and inline images come through. Check the tables first: column widths follow the document's own grid, and a wide grid is scaled to fit rather than being allowed to run off the page.

A simple form to send out for signing

Text, checkboxes typed as characters, indents and underlines are reproduced. Form fields built as Word content controls render as their current text, not as fillable PDF fields — this produces a document to print or sign by hand, not an interactive form.

A letter on headed paper

The header and footer from the document are drawn on every page, along with the page margins that position them. A different first-page header is reported, because only the default one is used.

A CV or resume

The usual case for wanting a PDF at all, and the one most sensitive to font metrics. A CV built to fit exactly one page in Word may run a few lines onto a second here; check the page count before sending it to an employer.

Before you use the result

Fonts are the honest ceiling. A .docx names its fonts but almost never carries them, and the ones it names — Calibri, Cambria, Segoe UI — are licensed system fonts a web page cannot read. Families are therefore mapped onto the standard PDF fonts: serif faces to Times, monospaced faces to Courier, everything else to Helvetica. Those metrics differ slightly from Word's, so a line can break in a different place even when every other property is honoured. This is the single largest source of visual difference.
The standard PDF fonts cover Latin text and Western European accents. Greek, Cyrillic, Arabic, Hebrew, CJK and most symbols and emoji fall outside them and are replaced with a question mark. The count of substituted characters is reported rather than left for you to find.
PNG and JPEG images are drawn. EMF, WMF, SVG and other formats a PDF cannot carry directly are named in a warning and left out, because converting them would mean guessing at a rasterisation the document never specified.
Text boxes, drawing canvases, floating shapes, embedded objects, charts and equations are not drawn. Footnotes keep their reference marks but the notes are not placed at the foot of the page, and field codes such as page numbers show their last cached value rather than being recalculated.
Vertically merged table cells and nested tables are reported rather than reproduced, and only the document's first section governs page setup, so a file with several differently sized sections is laid out on one page size throughout.
This is a visual reconstruction, not a certified rendering. Where a document is going to a court, a government portal or a client under contract, open the PDF and read it before you send it.

Questions about DOCX to PDF

Will it look exactly like Microsoft Word?

No, and any converter claiming otherwise without embedding the original fonts is overstating it. The layout is reconstructed closely — page size, margins, styles, spacing, alignment, lists, tables and images — but the fonts are mapped rather than reproduced, so line breaks can fall differently. Documents with text boxes, charts or equations will differ more.

Are my files uploaded?

No. The .docx is unzipped, parsed and rendered inside this browser tab, and the PDF is written to your device from there. There is no endpoint on this site that accepts a file, which is checkable in your browser's network tab while you convert.

Does it support old .doc files?

No. Older .doc files are not supported; open the file in a word processor and save as .docx first. The two share a name and nothing else: .doc is a binary compound file from before 2007, while .docx is a ZIP of XML. A .doc is refused by its file signature rather than being half-read.

Does it preserve images and tables?

Tables keep their rows, cells, column widths, borders and cell shading, and rows are kept whole across page breaks. Inline PNG and JPEG images are redrawn at the size the document specifies, scaled down if they are wider than the text column. Vertically merged cells, nested tables and image formats a PDF cannot carry are reported instead.

What can change during conversion?

Line breaks, because of font metrics; anything in a script the standard PDF fonts do not cover; text boxes, shapes, charts, equations and embedded objects, which are not drawn; footnote placement; and field codes, which show their cached value. Every one of these is listed as a warning on this page before you download, so nothing is dropped without being named.

DOCX to PDF