WORKFLOW GUIDE
Pandoc Markdown to PDF
Use Pandoc for templates, citations and repeatable builds. Use the browser converter for one Markdown document without installing a PDF toolchain.
Guide revised 20 October 2026 · Documentation and implementation comparison
COMMAND LINE
Start with Pandoc
pandoc input.md -o output.pdfInstall Pandoc and a PDF engine first. The default PDF route uses pdflatex; other engines have their own installation requirements.
IN YOUR BROWSER
Export one document
Paste Markdown or choose one .md file. Review the preview, choose page settings, then download and check the actual PDF.
No account. Local files are not uploaded for conversion.
The command is documented Pandoc usage, not a conversion measured for this guide. Neither route guarantees the layout of an untested document.
CHOOSE WHAT THE JOB NEEDS
Compare the workflows
Pandoc is a document converter that passes PDF creation to an installed engine. Our browser tool renders Markdown for preview and builds the PDF locally. The table compares available controls and implementation, not output quality from a matched test.
| What matters | Pandoc workflow | This browser converter |
|---|---|---|
| Setup | Install Pandoc and the chosen PDF engine. | Open the page and load its conversion assets. No toolchain installation. |
| Inputs and outputs | File and format options control conversion. This guide covers Markdown input and PDF output. | Paste Markdown, choose or drop one .md file, or fetch a public Markdown URL. Download PDF or HTML. |
| Layout controls | Templates, variables and options for the selected engine. | A4 or Letter; portrait or landscape; Normal, Narrow or Wide margins; font and size; optional page break before each H2. |
| Fonts | Font selection depends on the engine and its configuration. | The font control offers Arial, Georgia and Courier New. PDF export maps those labels to Helvetica, Times and Courier. |
| Tables and pagination | Check the table features and page rules of the selected writer and engine. | The current writer uses equal-width columns and moves rows to the next page when needed. It does not repeat table headers automatically; an oversized row needs review. |
| Table of contents | --toc adds a table of contents in supported outputs. | No built-in table-of-contents control. |
| Citations | --citeproc processes citations when bibliographic data is supplied; a CSL style can control formatting. | No Pandoc citation processor, bibliography workflow or CSL input. |
| Automation | Command-line options, scripts and filters can form a repeatable workflow. | Interactive single-document export. No supplied CLI, batch export, folder conversion or conversion-history feature. |
| Local processing | Installed Pandoc can process local files on your machine. Inspect any remote resources, filters and engine behaviour used by your workflow. | Local Markdown is processed in the tab without a document upload. URL loading, remote images, page delivery and analytics can make network requests. |
Choose by the work you need to repeat
Choose Pandoc when…
- Your output depends on a template, bibliography or engine-specific typography.
- You want to keep conversion commands in a build or script.
- You are prepared to install and maintain the required engine, fonts and packages.
Choose the browser when…
- You have one document to export without setting up a toolchain.
- The available page controls are enough for your layout.
- You can review the downloaded PDF before using or sharing it.
Both routes can process local documents. Local handling alone is not a reason to claim that one produces a better PDF.
PREVIEW IS NOT THE FINAL FILE
What the downloaded PDF can change
Our PDF writer rebuilds the preview content with jsPDF; it does not print the entire preview unchanged. These are boundaries found in the current source, not measurements of conversion accuracy.
Font appearance
The PDF uses the mapped PDF fonts listed above. The font-control label is not a promise that a font of that exact name is embedded. Review line wrapping in the PDF separately from the preview.
Links
The PDF writer adds link destinations as visible URLs alongside link text. It does not add clickable link annotations in the current implementation.
Tables and long blocks
Equal-width table columns may wrap differently from the preview. Table headers are not automatically repeated. A row or visual block taller than a printable page can exceed the available area; inspect the output instead of treating a completed export as proof of a fit.
Diagrams, formulas and code
The preview uses Mermaid, KaTeX and syntax highlighting. PDF export has its own text and image handling; those preview features are not a guarantee of identical PDF rendering. Check formulas, diagram labels, code spacing and symbols in the downloaded file.
For document-specific preparation, read the Markdown table guidance, Mermaid to PDF or LaTeX to PDF. These share the browser export workflow; they do not add Pandoc templates or citation processing.
This guide does not publish a matched Pandoc-versus-browser sample run. There is no measured speed winner, accuracy score, typography ranking or claim that either route handles every document. Pandoc results depend on the selected engine, version, fonts and options.
Local conversion and network requests
Pasted text and local .md files are previewed and exported in your browser without being uploaded to a conversion server. This is a file-handling statement, not a promise that the page makes no network requests.
- Load from a URL
- Your browser asks that address for Markdown. It retrieves content; it does not upload your local Markdown to that address. The source server can see the request, and its cross-origin rules may block it.
- Remote images
- A Markdown image URL can contact its host while the preview loads or the export attempts to embed it. An image may appear in the preview but fail to export if its host restricts cross-origin access.
- Page visits
- Site delivery and analytics have separate network requests. See the privacy policy for the data collected. Document contents and conversion results are not included in our analytics events.
Pandoc also supports local processing. A local command does not by itself establish that all resources stay offline: check the URLs, filters and engine options used in your own workflow.
Questions before you export
What if Pandoc cannot find a PDF engine?
Install the engine required by your command and make it available on your system path. The default route uses pdflatex. To choose a different installed engine, use --pdf-engine; an engine such as Typst or WeasyPrint has its own setup. Follow the official engine instructions.
Can I use my Pandoc template in the browser tool?
No. This browser workflow does not accept Pandoc templates, Lua filters, variables or CSL files. Its page controls are independent of Pandoc, and it does not require Pandoc to be installed.
Why does a public Markdown URL fail to load?
The browser requests the URL directly. A failed response, an unavailable file or the host's cross-origin restrictions can stop it. If you can access the Markdown yourself, save it as a local .md file or paste its text instead.
Can I trust a correct-looking preview?
Use it to check the Markdown structure, then inspect the downloaded PDF. Check page breaks, fonts, table rows, figures and symbols. A successful export reports completion, not a complete quality check.
Sources and next steps
Pandoc capability statements and the command above are based on the official user guide. Browser capability statements describe our current controls and implementation; they are not a side-by-side output benchmark.