The same address, www.republic.se/en/, fetched twice with curl and cut to the first twelve lines.
Without asking for a format you get HTML: the doctype, meta tags and the start of a script. With the
header Accept: text/markdown you get Markdown: the page's title, description and first heading.