Markdown to PDF, Word, HTML: The Conversion Playbook

Markdown is the format you write in. PDF, Word, and HTML are the formats other people receive. Somewhere between the two sits a conversion step that looks trivial and turns out to be full of small traps: a code block that bleeds off a PDF page, a Word file that arrives as flat text with no heading styles, an HTML export that leaks your YAML frontmatter at the top. The good news is that the same handful of tools cover all three targets, and once you know which tool maps to which job, converting Markdown stops being a guessing game.

This is the complete playbook for converting Markdown to the three formats people actually ask for — PDF, Word (.docx), and HTML — with the reliable methods for each, honest pros and cons, and copy-paste commands. If you only need PDF, we have a dedicated deep dive on the five ways to convert Markdown to PDF; this guide is the wider map across all three.

TL;DR — To convert Markdown to PDF, Word, and HTML, Pandoc is the one tool that does all three well (pandoc doc.md -o doc.pdf, -o doc.docx, -s -o doc.html) and reads your frontmatter as metadata. For a quick PDF of a single note, use your editor's export or print-to-PDF; for HTML fragments, any Markdown library works. The one thing that keeps every path open is keeping your source as portable .md — which is exactly what MDflow stores, with a one-click PDF export built in.

The one idea that makes all conversions make sense

Every Markdown converter does the same two-step underneath: it parses your Markdown into a structured document (headings, paragraphs, lists, a table here, a code block there), then renders that structure into a target format. The quality of a conversion comes down to how faithfully those two steps preserve your intent.

That is why Pandoc is in a category of its own: instead of translating Markdown directly to each output, it parses Markdown into a rich internal document model — an "abstract syntax tree" — and then renders that to PDF, Word, HTML, or any of forty-odd formats. One faithful parse, many faithful outputs. Most other tools are single-purpose wrappers around a browser engine (great for PDF and HTML, since browsers speak both natively) and simply have no path to a real .docx.

Keep that split in mind and the whole landscape falls into place:

Markdown to PDF

The fastest way to get a PDF is your editor's "Export to PDF" button; the highest-quality way is Pandoc. PDF is the most-requested target and the one with the most options, because two different engines can produce it — a Chromium browser print engine (easy CSS styling) or a LaTeX typesetting engine (book-grade pagination).

For the everyday case — one good-looking PDF of one document — reach for a built-in export (MDflow, Typora, VS Code, Obsidian all ship one) or just render the Markdown anywhere and press Cmd/Ctrl+P → Save as PDF. For a polished, repeatable, print-quality document, use Pandoc with a LaTeX engine:

pandoc document.md -o document.pdf --pdf-engine=xelatex

The --pdf-engine=xelatex flag matters: the default pdflatex cannot use system fonts, while xelatex (or lualatex) lets you set any installed font. The three things that reliably go wrong with Markdown-to-PDF — page breaks, fonts, and frontmatter — plus all five methods in detail are covered in the companion post, Markdown to PDF: The 5 Reliable Ways. The short version: browser engines break pages with CSS page-break-after: always, LaTeX breaks with a raw \newpage, and only tools that treat --- as metadata (Pandoc does) keep frontmatter out of the body.

Markdown to Word (.docx)

Pandoc is the best — and effectively only reliable — way to turn Markdown into a real, editable Word document:

pandoc document.md -o document.docx

This is the conversion people most often get wrong, because the tempting shortcuts all produce fakes. Copy-pasting rendered Markdown into Word loses the semantic heading structure; "printing" to Word gives you flat text; and browser-engine tools have no .docx path at all. Pandoc is different: it emits a native .docx with proper Word styles — Heading 1, Heading 2, List Bullet, a real table object — so the recipient can keep editing it in Word or Google Docs exactly as if a human had typed it there.

To match a corporate template or house style, hand Pandoc a reference document:

pandoc document.md --reference-doc=house-style.docx -o document.docx

Pandoc reads the styles, margins, fonts, headers, and footers out of house-style.docx and applies them to your output. (Generate a starter template to edit with pandoc -o custom-reference.docx --print-default-data-file reference.docx.) This is what makes Markdown viable in a Word-centric organization: authors write plain .md, and a single command produces a document indistinguishable from a hand-built one.

• Pros: genuine, editable .docx with semantic styles; template support for brand consistency; consumes YAML frontmatter as metadata; scriptable for batch and CI.
• Cons: you must install Pandoc; raw HTML embedded in your Markdown mostly does not survive the trip to Word (Word has no equivalent for arbitrary HTML), so stick to standard Markdown constructs for clean output; very complex layouts still belong in a real DTP tool.

Markdown to HTML

Converting Markdown to HTML is the easiest of the three, because HTML is the format Markdown was invented to generate — the real choice is whether you want a fragment or a full page.

A fragment is the inner HTML — <h1>, <p>, <table> — with no <html>/<head> wrapper, ideal for pasting into a CMS, a static-site template, or an email. Every Markdown library produces it, and you almost certainly already have one:

# Node
npx marked document.md > fragment.html

# Python
python -m markdown document.md > fragment.html

Libraries like markdown-it, marked, remark, and Python-Markdown power most of the web's Markdown rendering and give you the fragment directly.

A standalone page — a self-contained .html file you can open in a browser or host — is Pandoc's job:

pandoc document.md -s -o document.html

The -s (--standalone) flag wraps the output in a proper HTML document with <head> and <body>. To make it a single portable file with the styling and images baked in — nothing to link, nothing to break when you move it — inline everything:

pandoc document.md -s --embed-resources -c style.css -o document.html

--embed-resources converts linked CSS, scripts, and images into data: URIs so the whole page travels as one file; -c style.css points at the stylesheet to embed.

• Pros: trivial and dependency-light for fragments; Pandoc's standalone mode gives clean, portable pages; full control over styling via CSS; the natural on-ramp to publishing on the web.
• Cons: a bare fragment is unstyled until you add CSS; sanitize HTML if the Markdown came from an untrusted source (most libraries have a safe mode); "standalone" and "self-contained" are different flags — use --embed-resources when you actually need one portable file.

Which applications benefit most

Multi-format Markdown conversion earns its keep wherever one source has to reach several audiences:

1. Documentation teams — author docs in Markdown, publish to HTML for the web, and hand a .docx to legal or a PDF to a client, all from the same file.
2. Reports and deliverables — write once, ship a print-quality PDF to the customer and an editable Word version to the internal reviewer.
3. Academic and technical writing — Markdown plus Pandoc covers PDF (via LaTeX) for submission and HTML for a preprint, with citations intact.
4. Static sites and blogs — Markdown-to-HTML is the entire content pipeline behind Jekyll, Hugo, Astro, Next.js, and every other static generator.
5. AI and automation pipelines — an agent or CI job that produces Markdown can fan it out to whatever format the destination system expects, deterministically.

How MDflow fits

MDflow is a markdown workspace in the browser, and it is built on exactly the principle this playbook rests on: keep the source portable, and every conversion path stays open.

What lines up today:

• One-click PDF export. Every document has an Export to PDF action — in the editor and in the read-only reader for shared documents — that renders your Markdown through a stylesheet tuned for paper (sensible margins, clean styling for headings, tables, and code) and hands off to your browser's Save as PDF. No frontmatter leaks, nothing to install. It is the "editor export" method from the PDF guide, done for you.
• Your .md, always. Because MDflow is markdown-native — your document is plain .md, not a proprietary blob — you can download the original at any time and pipe it straight into Pandoc for Word or HTML: pandoc doc.md -o doc.docx, pandoc doc.md -s -o doc.html. The portable source is the on-ramp to every method in this guide.
• Programmatic access. Fetch the raw .md twin of any shared document over the HTTP API, or pull it through the MCP server, and run Pandoc on the other side. An AI agent or a build step can convert on demand without ever touching a browser.
• Whole-workspace export. Export an entire workspace as a self-contained archive of .md files that renders offline in Obsidian, VS Code, or GitHub — a clean handoff into any downstream conversion pipeline.

To be honest about the boundary: MDflow itself converts to PDF in one click and gives you the portable .md for everything else — it does not render Word or HTML in the app. For those, the workflow is deliberately simple: take the .md MDflow hands you and run one Pandoc command.

Where we are headed is direction, not a dated commitment: richer export themes and page options, and eventually more one-click targets so more of this playbook happens without leaving the app. The throughline is unchanged — your content stays plain Markdown, so it is never trapped in one format.

The bottom line

There is no single best converter — there is a best tool per target. For PDF, use an editor export or print-to-PDF for speed, Pandoc for print quality. For Word, use Pandoc with a --reference-doc template — it is the only path to a genuinely editable .docx. For HTML, use any Markdown library for a fragment, or Pandoc's -s --embed-resources for a self-contained page. And if you want one tool for all three, that tool is Pandoc, because it parses your Markdown once and renders it faithfully everywhere.

The constant underneath all of it: the more portable your source, the more paths stay open. Keep it in plain Markdown, and PDF, Word, and HTML are all one command away.

If your Markdown already lives in MDflow, PDF is one click and the source is always yours to take anywhere.

Start free · Connect an AI agent · Read the API docs

Frequently asked questions

How do I convert Markdown to PDF, Word, and HTML with one tool?

Pandoc converts Markdown to all three from the command line: pandoc doc.md -o doc.pdf for PDF, pandoc doc.md -o doc.docx for Word, and pandoc doc.md -s -o doc.html for standalone HTML. It is the single most capable Markdown converter, reads your YAML frontmatter as document metadata, and supports custom templates for each format. The trade-off is installing it (plus a TeX distribution if you want PDF).

What is the best way to convert Markdown to Word (.docx)?

Pandoc is the best Markdown-to-Word converter: run pandoc doc.md -o doc.docx to get a real, editable .docx. To match a house style, pass --reference-doc=template.docx and Pandoc applies that document's styles, margins, fonts, headers, and footers. This produces a native Word file with proper heading styles — not a screenshot or a print — so anyone can keep editing it in Word or Google Docs.

How do I convert Markdown to HTML?

For a full standalone web page, run pandoc doc.md -s -o doc.html, and add --embed-resources -c style.css to inline a stylesheet and images into a single self-contained file. If you only need the HTML fragment (to paste into a CMS or template), most Markdown libraries do it: markdown-it, marked, remark, or Python-Markdown all output the inner HTML without a full page wrapper.

Does converting Markdown lose my formatting?

Standard formatting — headings, lists, tables, links, code blocks, bold and italic, images — survives conversion to PDF, Word, and HTML because it maps to native constructs in each format. What can be lost is anything non-standard: raw HTML embeds may not carry into Word, and extended syntaxes like footnotes or math need the right converter flags. Pandoc preserves the most because it parses Markdown into a rich internal document model first.

Can MDflow export to PDF, Word, and HTML?

MDflow has a built-in one-click print-ready PDF export in the editor and the shared-document reader, and it lets you download the original .md at any time. For Word and HTML, take that portable .md and run it through Pandoc — pandoc doc.md -o doc.docx or pandoc doc.md -s -o doc.html. Because MDflow stores plain Markdown rather than a proprietary format, every conversion path stays open to you.

Further reading

• Pandoc — User's Guide (--reference-doc, --embed-resources, -s)
• Pandoc — Creating a Word template (reference-doc)
• markdown-it — Markdown parser on GitHub
• MDflow — Markdown to PDF: The 5 Reliable Ways · Markdown for AI agents · API documentation · FAQ