Paste Markdown or drop a .md file, set the page size and margins, and download a PDF that matches the preview exactly. Tables get borders, code blocks get a background and blockquotes get a rule — without that styling a converted Markdown table renders as bare text and looks like the conversion failed. The output is rendered from the preview, which means it is pixel-accurate and its text is an image rather than selectable. If you need searchable text, convert to Word instead and export from there. Nothing is uploaded, and raw HTML in your Markdown is sanitised before it is rendered.
| Word | ||
|---|---|---|
| Matches preview exactly | Yes | No — reflows |
| Text selectable | No | Yes |
| Editable afterwards | No | Yes |
| Best for | Printing and sharing | Further editing |
No. Parsing and rendering both happen in your browser.
Because the page is rasterised to match the preview exactly. That is the trade for pixel fidelity. The Markdown to Word tool produces real editable text, and Word exports a searchable PDF from it.
Yes, with borders and a bold header row. GitHub-style pipe tables work as written.
It is sanitised first. Markdown permits raw HTML, so without that a document containing a script tag would run code in this page. Scripts and frames are removed; ordinary markup renders.
Not directly. Content flows and breaks where the page ends. Adding blank lines before a heading nudges it, but there is no explicit page-break control.
Images referenced by URL are rendered if they load and permit cross-origin use. Local file paths will not resolve, since the browser cannot read your filesystem.
Markdown is parsed with marked, sanitised with DOMPurify, laid out off-screen at the page's content width so the line breaks in the PDF are the ones that width actually produces, then rasterised and sliced across pages. Sanitising matters even though the input is Markdown: the format permits raw HTML, so a document containing a script tag would otherwise execute inside this page. The resulting PDF matches the preview exactly and its text is an image — unsearchable and unselectable — which is the honest trade for that fidelity, and the page says so and points at the Word converter for anyone who needs real text. Each page takes its own slice of the tall canvas rather than the whole bitmap offset upward, which stops a long document embedding the entire image once per page and producing a file several times larger than it should be. Tables, code blocks and blockquotes carry explicit styling, because the browser's defaults render a Markdown table without borders, which looks like a failed conversion rather than a styling choice.