Word to PDF
Convert .docx files to PDF with a rendering mode that matches your document.
100% private100% private
Convert a .docx file into a PDF that preserves formatting, fonts, headers, footers, and embedded images. Useful when an upload portal demands PDF and you only have Word, or when you want a fixed-layout final document that won't reflow on someone else's machine. The conversion runs locally in your browser on top of pdf-lib, which means proprietary drafts and confidential reports are not round-tripped through a third-party converter just to change extensions.
Processed on your machine and only your machine. No files are sent to any server.
BetaNext steps
Runs right inside your browser tab. No uploads. Your files stay private.
How Browser-Based Word to PDF Conversion Works — Two Pipelines, Trade-offs
Word to PDF converts .docx files into PDF in the page itself, never on a server, with no server involvement. The tool offers two conversion engines you choose between with the toggle above — Direct Parse and HTML Render — each with different trade-offs. Both keep the file in tab memory the whole time — nothing is uploaded.
The Direct Parse engine reads the .docx ZIP archive (DOCX is XML-in-a-ZIP), parses the OOXML structure with a custom parser, runs deterministic layout, and writes PDF primitives via pdf-lib. This route preserves text as text (selectable, searchable, copy-paste-able) and produces a small, vector-clean PDF, roughly 50–500 KB for a typical document. Fonts are mapped to the standard PDF font families — Calibri and Arial render as Helvetica, and Times New Roman is preserved — so the shapes are close but not pixel-identical to the original typeface. It is the better choice when the document uses standard paragraphs, headings, lists, simple tables, and inline images. Direct Parse deliberately rejects documents that use features it cannot reproduce faithfully (such as hyperlinks, complex field codes, equations, charts, SmartArt, or embedded fonts) with a clear message rather than producing a wrong-looking PDF — switch to HTML Render for those.
The HTML Render engine uses mammoth (DOCX → semantic HTML), renders the HTML inside an isolated iframe at 816×1056 pixels (US Letter at 96 DPI), captures each page with html2canvas, and embeds the resulting PNG images into a PDF using pdf-lib. Choose this engine for documents Direct Parse rejects, or when you want a closer visual match to how the document looks on screen. The output is a raster PDF — every page is an image, so text is no longer selectable and file size is significantly larger.
The conversion tries to preserve heading levels, paragraph styles, bold/italic/underline runs, ordered and unordered lists, simple tables, and inline images (PNG/JPEG embedded as PDF XObjects). It also detects title-like centered content via heuristics and applies the appropriate alignment. Note that hyperlinks are not preserved as clickable links — Direct Parse rejects documents containing them, and HTML Render flattens them into the page image.
Realistic limits: the .doc legacy binary format (Word 97–2003) is not supported — only .docx. Documents with complex floating shapes, equation editor (OMath), embedded charts, footnotes that span pages, multi-column layouts, headers/footers with field codes (PAGE OF NUMPAGES), and SmartArt are rejected by Direct Parse and may render imperfectly under HTML Render. For those documents, the most reliable conversion is still Microsoft Word's built-in Save As PDF or LibreOffice in headless mode.
Image fidelity in the direct pipeline matches the source — embedded images are pulled from the .docx ZIP and re-embedded into the PDF without re-encoding, so quality is preserved. The HTML pipeline rasterizes everything at 96 DPI, which is fine for screen reading but soft when zoomed or printed at full Letter size.
Browser memory is the practical limit. .docx files up to ~30 MB and 200 pages convert comfortably. Documents with hundreds of high-resolution embedded images can pressure the tab's heap during the html2canvas pass; in those cases, the direct pipeline is dramatically lighter and usually succeeds where the raster pipeline fails.
Common Use Cases
01
Resume and CV submission
Convert a Word resume to PDF so layout, fonts, and spacing render identically on every recruiter's machine, regardless of Word version.
02
Client-ready proposals
Send polished PDFs that don't require Microsoft Word to open and don't accidentally show tracked changes or comments.
03
Long-term archival
Convert documents to PDF/A-style stable output that opens reliably for years, even after Word format changes or new template defaults.
04
Print-ready handouts
Generate PDFs that print services accept directly without asking for the source .docx and the fonts used.
Frequently Asked Questions
.docx only — the XML-based format introduced in Word 2007. The legacy binary .doc format (Word 97–2003) is not supported. Convert .doc to .docx in Word or LibreOffice first.
Yes when you use the Direct Parse engine — text remains as selectable, searchable, copy-paste-able PDF text. The HTML Render engine rasterizes each page to a PNG image and loses selectable text. You pick the engine with the toggle at the top of the tool; the tool does not switch between them automatically.
No. The .docx is parsed and converted without ever leaving your tab — mammoth, html2canvas, the custom OOXML parser, and pdf-lib all run in the tab. The file never reaches a server.
Standard tables, ordered/unordered lists, and inline PNG/JPEG images are preserved by the Direct Parse engine. Documents with complex floating shapes, equation editor content, charts, SmartArt, or multi-column layouts are rejected by Direct Parse — switch to HTML Render, which captures the document as a page image and handles a wider range of layouts at the cost of selectable text.
Currently one file at a time. The direct pipeline is fast (typically under a second per page), so batch conversion is straightforward by repeating the upload for each file.
The Direct Parse engine maps fonts to the standard PDF font families rather than embedding the original typeface: Calibri and Arial render as Helvetica, Times New Roman is preserved, and Courier New maps to Courier. Shapes are close but not pixel-identical. The HTML Render engine uses whatever fonts your browser has installed, so for a closer match install the document's fonts in your OS, or use desktop Word for pixel-perfect typography.
The Direct Parse engine renders headers and footers as static content and shows a conversion note that dynamic page numbers and alternating content aren't reproduced. Documents that rely on complex field codes (such as PAGE / NUMPAGES counters) or content controls are rejected by Direct Parse — use HTML Render, which captures whatever your browser renders.
Use HTML Render when Direct Parse rejects your document — it analyzes the file first and refuses anything with features it can't reproduce faithfully (math equations, complex floating layouts, SmartArt, charts, hyperlinks, embedded fonts), reporting exactly what it found. HTML Render handles those by capturing the page as an image instead. The trade-off is image-only output with no selectable text and a larger file.
Direct Parse output has selectable text and can be annotated with PDF Editor. HTML Render output is image-only — to make it editable or searchable again, run it through PDF OCR.
Direct Parse rejects documents that still contain tracked changes or comments. HTML Render converts the document text but does not reproduce change bars or comment bubbles. Either way, the cleanest result comes from accepting all changes and removing all comments in Word before converting.
Step-by-step guide
How to convert Word to PDF
Walk through every step with screenshots, format-specific tips, and the platform-by-platform limits you need to know.
Advertisement
What's next?