Convert HTML to Markdown

The words, and nothing else. Markdown stores characters and line breaks, so bold, italics, headings, fonts, colours, tables, images, headers and footers are all discarded — not degraded, removed. That is the correct outcome when you want the text for indexing, diffing or scripting, and the wrong one if the formatting carried meaning. The long form is what Jekyll, Hugo and several documentation pipelines look for by default, so it fits straight into those builds. It inherits every Markdown ambiguity — dialect differences around tables, footnotes and raw HTML — with no benefit over .md.

Stylesheets are resolved and applied before conversion rather than carried across, because only HTML has a separate presentation layer. Images referenced by relative path are fetched and embedded so the output is self-contained; images behind authentication or JavaScript rendering will not be. Absolute links stay clickable. HTML expresses content as a tag tree, with presentation delegated to CSS and any images referenced rather than embedded.

CommonMark plus the GitHub Flavored Markdown extensions, which is the combination almost every current tool understands. That matters because Markdown was never standardised in one place: tables, footnotes, strikethrough and raw HTML behave differently between dialects, and those are exactly the constructs that silently change meaning when a document moves between renderers.

Yes — basic HTML → Markdown conversion is free with no registration. Premium adds higher per-file size limits, more parallel batch workers, and removes ads.

Free accounts: up to 100 MB per HTML file. Premium: typically 1 GB per file, depending on the HTML source type. Very large files are split into chunks server-side when the Markdown format allows.

Conversion preserves all information that both HTML and Markdown can represent. Format-specific features that the Markdown format doesn't support (e.g. vector data in raster output, formulas in plain-text output) are rasterized or stripped — see the format-specific notes on the result page.

Yes — drop multiple HTML files into the upload zone and they are queued in parallel. Premium has more parallel workers; on a 50-file batch this is the difference between minutes and seconds.

Yes — the UI is responsive and the conversion pipeline runs server-side, so mobile devices upload-then-download just like desktops. iOS and Android browser file pickers (camera roll, Files app) work end-to-end.

Chrome, Firefox, Safari (15+), Edge, and Opera. The converter uses standard `<input type="file">` upload — no WebAssembly or PWA dependencies — so even older browsers work.

Yes — uploaded HTML files are processed in isolated workers and deleted within minutes. We never view, retain, or share content. See /privacy/ for the data-retention window.

Disable any pop-up blocker, then click the download button manually. The result link stays active for 24 hours, so if your network fails mid-download, you can re-trigger from the result page without re-uploading the HTML.