Convert Word to PDF — fonts, tables, tracked changes, and PDF/A

The short version

1. Have Word installed? File → Save As → PDF. Highest fidelity; embeds fonts; handles tracked changes explicitly.
2. Doc lives in Google Docs? File → Download → PDF Document. Good; strips tracked changes; needs a Google account.
3. Don't want to install or sign in? Drop the .docx into our converter. Anonymous, embeds common fonts, preserves tables and images, no watermark.
4. Library or government submission? You probably need PDF/A. Word does it ("ISO 19005-1 compliant" checkbox); most online converters don't.
5. Rare fonts? Embed them or accept substitution. Substitution is the #1 reason a PDF looks "almost right but the line breaks moved."
6. PDF came out 30 MB? Chain through our compressor — image-heavy DOCX balloons the output.

Why conversion isn't always faithful

DOCX is a zipped XML format that describes paragraphs, runs, tables, and references to images and fonts. PDF is a fixed-layout format describing drawing commands: put this glyph at this coordinate, stroke this path, embed this image. The converter has to bridge the two — decide where each line breaks, which font renders each run, how to flow a multi-page table, what to do with a comment balloon.

Small differences produce very visible downstream effects. A font substitution that changes character width by 2% shifts line breaks, which shifts paragraphs, which pushes a heading onto the next page. A table renderer that doesn't understand nested cells merges them wrong. A converter that ignores the tracked-changes state bakes every red strikethrough into the final PDF — sometimes what you want, usually not. Pick the path that matches what your document actually contains.

Path 1: Word's built-in Save as PDF (highest fidelity)

If Word is already installed, its exporter is the reference implementation — same layout engine that rendered the doc on screen, so the PDF comes out pixel-identical. The downside is needing a Microsoft 365 license ($6.99+/mo) or a perpetual Office install. File → Save As → PDF on Windows, or File → Save As → File Format: PDF on macOS. Click the Options button.

The checkboxes worth knowing:

• ISO 19005-1 compliant (PDF/A). Check it for library, archive, or government submissions. Rejects unembedded fonts and dynamic features.
• Optimize for Standard vs Minimum size. Minimum downsamples images to ~150 DPI; Standard keeps source resolution. Email: Minimum. Print: Standard.
• Document structure tags for accessibility. Keep checked. Required for Section 508 (US) and EN 301 549 (EU) compliance.
• Print Markup. If checked, tracked changes and comment balloons bake into the PDF. Uncheck for a clean final. Easy to miss; lots of red-ink-in-final-PDF accidents start here.

If you have Word, use it. Tracked-changes behavior is explicit, PDF/A is one checkbox away, and the PDF matches your on-screen preview exactly. The rest of the paths are for when you don't.

Path 2: Google Docs PDF export (good, needs an account)

Drag the .docx into Drive (or File → Open). Google silently converts to its own format, then File → Download → PDF Document produces the PDF. Works in any browser; no install.

It handles common fonts, images, tables, and headers/footers cleanly. It's free. Where it loses:

• Account required, and the file routes through Google Drive. For PII or client-confidential content that isn't already in Google's ecosystem, that's a consent question worth asking.
• Tracked changes and comments don't survive the DOCX-to-Docs import cleanly — they convert to Suggestions and Comments, and PDF export strips both by default. If tracked changes are the whole point (legal redlines), this path silently discards them.
• Complex formatting (nested tables, multi-column layouts, field codes, old Equation Editor equations) can reflow during the DOCX reinterpretation. Skim before exporting.
• No PDF/A export. Not usable for archival.
• Uncommon fonts substitute from Google's web-font catalog. Arial, Times New Roman, Calibri, Georgia, Helvetica survive; specialty brand fonts usually don't.

Path 3: Online converter (anonymous, no install — our angle)

This is where we fit. You don't have Word, you don't want to hand Google your document, and you need the PDF in the next thirty seconds. What a good online converter has to do: embed fonts, preserve tables and lists, keep images at a sensible resolution, resolve headers/footers/page numbers, handle tracked changes deliberately (not silently garble them), and not watermark the output.

Our DOCX to PDF does all six. We render via LibreOffice headless server-side, embed fonts, preserve images at source resolution (with a sanity cap), and accept tracked changes as final by default. The output is a plain, unstamped PDF. Where online converters lose to Word: PDF/A certification, form-fillable fields, and exotic fonts you haven't uploaded.

PDF/A — the version for archives

PDF/A is a stricter subset of PDF designed for long-term archival. The spec (ISO 19005) forbids features that depend on external state: unembedded fonts, expiring hyperlinks, JavaScript, encryption. A PDF/A file is guaranteed to render the same in 2026, 2046, and 2066.

Who asks for it: academic libraries (thesis submissions), US federal records (NARA), EU government portals, many courts, and ISO-audited corporate records-retention workflows.

Getting PDF/A:

• Word: Save As → PDF → Options → check "ISO 19005-1 compliant (PDF/A)."
• LibreOffice: File → Export As → Export As PDF → "Archive (PDF/A, ISO 19005)." Pick PDF/A-1a or PDF/A-2b depending on the requester.
• Acrobat Pro: File → Save As Other → Archivable PDF. $22.99/mo subscription.
• Online converters: most don't expose it. Ours is adding a toggle in a future release; for strict submissions use Word or LibreOffice today.

Regular PDF looks identical to PDF/A on screen but upload portals reject it with cryptic errors. Read the submission requirements before picking a converter.

Form-fillable PDFs — Word can't, Acrobat can

If you need a PDF with fillable text fields, checkboxes, or signature lines, Word's Save As PDF will not produce one — the form fields in your DOCX (gray shaded boxes) convert to static text. What actually creates a fillable PDF: Adobe Acrobat Pro ($22.99/mo, Prepare Form tool), Nitro PDF Pro ($179 one-time), PDFescape online (free tier), or LibreOffice Draw (free, Form Control toolbar, clunkier UI). The standard workflow is to produce a static PDF first (any path), open it in one of those, add fields, save. Don't try to add fields in the DOCX — they won't survive conversion.

Font substitution — the #1 reason PDFs look wrong

A DOCX references font names ("Calibri Light", "Futura PT"); the renderer has to find a matching file on disk. When it can't, it substitutes — picks the closest match. Substitution works most of the time, but it shifts character widths, which shifts line breaks, which shifts pagination. A 12-page doc becomes 13, a heading slides to the wrong page, a centered title reflows ragged.

The fix is embedding. Embedded fonts live inside the PDF — every reader renders identically regardless of what's installed. Every modern converter embeds by default; the issue is when the converter doesn't have a particular font to embed. Word on your machine embeds any font you have installed. Google Docs has a few hundred web fonts. We ship the full Google Fonts catalog plus the Microsoft core fonts (Arial, Calibri, Cambria, Consolas, Georgia, Times New Roman, Verdana). Bare online converters often run a Linux server with maybe ten fonts — anything exotic gets substituted aggressively.

For brand fonts or commissioned typefaces, the only safe paths are (a) Word where the font is installed, or (b) a converter that accepts the font file alongside the DOCX. Otherwise accept substitution and re-check the layout after conversion.

Tables, layouts, images, page numbers

Where converters diverge most visibly:

• Tables. Simple rectangular tables survive every converter. Nested tables, merged cells spanning three or more rows, tables that cross page boundaries, and auto-fit widths are where things break. Test with a sample before committing to a batch.
• Multi-column layouts. Word's Page Layout → Columns (2 or 3 columns, used for newsletters or academic papers) renders correctly in Word and LibreOffice. Some online converters flatten it to a single column.
• Embedded images. JPEGs survive intact. PNGs with transparency sometimes flatten to white. SVGs pasted in newer Word re-rasterize to 150 DPI PNG — the curves you thought were vector become bitmaps. Export and combine in Acrobat if the vector matters.
• Headers, footers, page numbers. These are template-driven Word fields (PAGE, NUMPAGES) that resolve at render time. A good converter resolves every field before embedding, so "7 of 23" shows up in the PDF. A bad converter drops the field, leaving blank headers. Spot-check first and last pages.

Tracked changes and comments

Word docs often carry review markup — tracked changes (red strikethrough and underline) and comment balloons. The converter has to pick a default: accept all changes, drop all comments (clean final PDF — our default, right for contracts and resumes), show markup (red stays visible, comments in a margin column — right for legal redlines), or reject all changes (niche; usually you'd do this in Word first).

The trap: most online converters don't ask, they just pick. A "show markup" converter bakes ignored red underlines into the final PDF. A silent-accept converter loses the audit trail when you needed it. Open the result and check page 1 before sending.

Batch workflow — 50 invoices in one pass

Drop the folder into our converter. Set global options (accept tracked changes, embed fonts) once, applied to every file. Server-side LibreOffice renders in parallel — 50 docs finish in a couple of minutes. Download a ZIP with filenames preserved. Command-line users can reproduce this locally with libreoffice --headless --convert-to pdf *.docx; that's effectively what we run on the back end.

How the options compare (honestly)

Every path has a right use case. Scoresheet:

Honest summary: if Word is installed, use it. If you're already living in Google Docs, use that. If you need to convert anonymously, quickly, and without installing anything — we're the shortest path. For PDF/A archival submissions and form-fillable PDFs, Word or Acrobat are the right answer; we don't pretend otherwise.

Works well / doesn't work

Works well

• Clean documents with common fonts (Calibri, Arial, Times New Roman, Georgia)
• Standard tables, bullet and numbered lists, headings
• Headers, footers, and page numbers (Word field codes resolve correctly)
• Embedded JPEGs and PNGs at reasonable resolution
• Documents with tracked changes you're ready to accept as final
• Batches of 10-500 similar documents (invoices, contracts, reports)

Doesn't work (well) yet

• PDF/A archival submissions — use Word, LibreOffice, or Acrobat for now
• Form-fillable PDFs — Acrobat or LibreOffice Draw are the right tools
• Proprietary brand fonts you haven't uploaded — will substitute
• Very old Equation Editor 3.0 equations — often garble; switch to the new equation tool in Word first
• Legal-redline PDFs where tracked changes must stay visible — set Word to "Print Markup" and export from Word

Tips for the best result

• Accept or reject tracked changes before converting unless you specifically want markup in the PDF.
• Compress images inside Word before export (File → Info → Compress Pictures → 150 ppi for email), or convert first and chain through our compressor.
• Swap exotic fonts for common equivalents if brand fidelity doesn't matter — avoids substitution debugging.
• Print-preview in Word first. If the Word preview looks wrong, the PDF will too.
• For government submissions, read the portal's requirements page. PDF/A-1, PDF/A-2, tagged PDF, and bookmark rules vary.
• For batches, convert 3 samples first. Check pagination, fonts, and tables before committing.
• Going the other way? Our PDF to Word extracts editable text from a PDF. Clean for text PDFs; best-effort for scans.

Common questions

Will the PDF look identical to my Word document?

If Word did the conversion, yes — same layout engine, pixel-identical. If an online converter or Google Docs did it, "visually very close" is the honest answer. Common fonts and simple layouts survive 99% of the time; nested tables, field codes, and specialty fonts are where differences creep in. Open the PDF and spot-check pages 1, middle, and last before sending.

Why is my converted PDF so large?

Embedded images at original resolution. A Word doc with ten full-page photos from a phone can land at 40+ MB. In Word: File → Info → Compress Pictures → 150 ppi ("Email") or 220 ppi ("Print"). Or convert first and run the result through our PDF compressor — image re-encoding typically cuts 70-80% off a photo-heavy doc.

Does your converter keep tracked changes?

By default, no — we accept all tracked changes as final and drop comment balloons. That's right for 90% of cases (contracts going out for signature, resumes, reports). If you need the markup visible in the PDF, open the doc in Word, set View → Markup → All Markup, and use Word's Save As PDF with "Print Markup" checked.

Can I convert DOCX to PDF/A for my library submission?

Today, our converter doesn't emit PDF/A — use Word (Save As PDF → Options → ISO 19005-1 compliant) or LibreOffice (Export As PDF → Archive). We're adding a PDF/A toggle in a future release; for now, strict archival workflows need Word or LibreOffice.

Can I create a form-fillable PDF from Word?

Not directly — Word's Save As PDF produces a static PDF, even if your DOCX has form fields. Produce the static PDF first (any path), then open it in Adobe Acrobat Pro (Prepare Form tool) or LibreOffice Draw (Form Control toolbar) to add fillable fields. PDFescape online is a free option for light form work.

My fonts look wrong in the PDF — why?

Font substitution. Your DOCX references a font the converter didn't have; it picked the closest match. Fix: use a converter that has your font (Word locally is safest) or swap to a common font (Calibri, Arial, Times New Roman, Georgia) that every converter has.

Are my files private on your tool?

The DOCX is uploaded for rendering (LibreOffice headless can't run in a browser tab), processed, and deleted within an hour. We don't log content, train on files, or require an account. For the highest-confidentiality work (legal, medical), Word's local Save As PDF never leaves your machine and is the safer choice.

Ready?

DOCX to PDF →. Drop a .docx, get a clean PDF with embedded fonts, preserved tables and images, and honest handling for tracked changes. Free, no watermark, no sign-up. If the result is bigger than you want, chain through our PDF compressor. If you're going the other direction — PDF in hand, DOCX needed — our PDF to Word is the reverse path. And if you need PDF/A, form fields, or proprietary-font embedding, use Word or Acrobat locally; we'll tell you so up front rather than silently produce the wrong thing.