Markdown to PDF: 3 routes that keep the formatting

All 3 run on your own machine (the print route only needs the rendered page). Which one fits depends on how often you will make this PDF again: once, use the browser; now and then, VS Code; for every release of a manual, pandoc in a script.

getPDF has no Markdown to PDF converter; it converts the other way (PDF to Markdown). Where it helps is after the PDF exists: page numbers for a handout, several chapters merged into 1 file.

Page numbersNumbers, header and footer, on turned pages too. Free, runs on your device.Open

The 3 routes side by side

	Browser print	VS Code, Markdown PDF extension	pandoc
Install	Nothing	VS Code and 1 extension	pandoc and a LaTeX engine
Look	The site’s or preview’s style	GitHub-like, your own CSS possible	LaTeX typesetting, or your template
Code highlighting	If the renderer highlights (GitHub does)	Yes, highlight.js, on by default	Yes, several styles
Table of contents	No	No	--toc
Links in the PDF	Usually clickable	Clickable	Clickable
Page breaks you control	CSS, where you can add it	<div class="page"></div>	\newpage, or a chapter per top heading
Good for	A one-off	Notes and READMEs you export now and then	Anything you regenerate

Route 1: print the rendered page

No install at all. You print what a renderer already drew.

1. Open the rendered file. On GitHub or GitLab, open the .md file in the repository; the site renders it. For a local file, use your editor’s preview, or any viewer that opens it in a browser tab.
2. Press Ctrl+P (Cmd+P on a Mac).
3. Set the destination to Save as PDF. In Chrome and Edge it is the Destination list; on a Mac, the PDF menu at the bottom of the dialog.
4. Open More settings: paper size A4 or Letter, margins, and untick Headers and footers if you do not want the URL and date on every page.
5. Save.

The catch: the PDF looks like the website, including its column width and whatever navigation the site lets print. Look at the preview in the dialog before you save. Fine for a quick copy, not for a handout.

Route 2: VS Code with the Markdown PDF extension

VS Code’s own Markdown preview (Ctrl+Shift+V) shows the rendered file but cannot print or export it. The extension most people use for that is Markdown PDF by yzane (version 2.2.0, over 4.2 million installs on the Visual Studio Marketplace, checked on 11 October 2026).

1. Install it: Extensions panel, search Markdown PDF, Install.
2. Open your .md file.
3. Open the command palette (Ctrl+Shift+P, Cmd+Shift+P on a Mac) and run Markdown PDF: Export (pdf).
4. The PDF lands next to the .md file, with the same name.

What it does, from its Marketplace page: it renders through a Chromium-based browser (an installed Chrome or Edge, or one it downloads on first use), the paper size is A4 by default (setting markdown-pdf.format), the margins are set with markdown-pdf.margin.top and its siblings, and code is highlighted with highlight.js unless you turn markdown-pdf.highlight off. The extension also exports HTML, PNG and JPEG from the same command list.

Route 3: pandoc, for anything you will regenerate

pandoc is the free command-line converter many technical writers build their documents with. One line gives an A4 PDF with a table of contents:

pandoc notes.md -o notes.pdf --toc -V papersize=a4 -V geometry:margin=2cm

The parts, as the pandoc manual describes them (pandoc.org, checked on 11 October 2026):

• --toc adds a generated table of contents from your headings.
• -V papersize=a4 sets the paper; letter is the other common value.
• -V geometry:margin=2cm passes the margin to LaTeX’s geometry package; repeat -V geometry:... for more options.
• -V fontsize=11pt sets the body size; the standard classes allow 10pt, 11pt and 12pt.

pandoc writes PDFs through a PDF engine, LaTeX by default, so it needs a TeX distribution (TeX Live, MiKTeX) installed. For your own fonts, or text in Greek, Cyrillic or other scripts the default engine struggles with, switch to XeLaTeX and name a font you have installed:

pandoc notes.md -o notes.pdf --toc --pdf-engine=xelatex -V mainfont="Liberation Serif" -V papersize=a4 -V geometry:margin=2cm

These commands follow the manual’s documented options; we did not run them for this page, because the laptop it was written on has no LaTeX install. Run them once on a short file before you build a 50-page manual on them.

How do I control page breaks?

Markdown has no idea of a page. A heading that lands at the bottom of page 3 with its text on page 4 is the commonest complaint. Each route has its own fix:

• pandoc: a line with only \newpage breaks the page (LaTeX passes it through). To start every top-level heading on a fresh page, make them chapters: add -V documentclass=report --top-level-division=chapter.
• Markdown PDF in VS Code: put <div class="page"></div> on its own line where the break goes; the extension documents that exact snippet.
• Browser print: if you control the CSS (your own preview or a static site), this rule starts each top heading on a new page:

h1 { break-before: page; }
pre, table, img { break-inside: avoid; }

The second line keeps a code block, table or picture from being cut in half across 2 pages, where the browser can.

After the PDF exists: number, merge, shrink

A manual written as 1 Markdown file per chapter becomes several PDFs. Three jobs follow, and getPDF does each in your browser with nothing uploaded:

1. Merge the chapters with Merge PDF: order them with Up and Down, and keep Bookmark per file on, so the result has 1 bookmark per chapter, with that chapter’s own bookmarks (pandoc makes them from your headings) nested under it.
2. Number the pages with Add page numbers: pick where (Bottom centre is the default) and the format, 1 or Page 1. Numbers are added as real text.
3. Compress it with Compress PDF if screenshots made it heavy. A PDF of text alone is usually small already, and the tool says so and returns it unchanged.

What these routes cannot do

The honest part:

• None of them is WYSIWYG. You see the layout only after you export. Keep a short test file for checking styles.
• Browser print is at the mercy of the site’s CSS. You cannot add a page break to GitHub’s rendering.
• pandoc’s LaTeX output looks like LaTeX. Making it look like your company’s template means writing or finding a template; that is a project, not a flag.
• Emoji and rare scripts need a font that has them, on every route.
• getPDF does not convert Markdown to PDF. We name no converter site here, because every route above already runs on your machine and keeps the file there. If you went the other way and want a PDF back as Markdown, PDF to Markdown does that on your device, and turning PDF pages into Markdown notes shows what it keeps.