What is an Accept-Charset header?
Tells the server which character encodings the client can understand, such as UTF-8.
Accept-Charset let older clients specify which text encodings they could render, back when a website couldn't just assume UTF-8 support everywhere. It's essentially obsolete today - UTF-8 has won so thoroughly that modern browsers don't even send it - so you're most likely to encounter it reading through old server logs or legacy client code rather than in anything currently worth monitoring.
Common use case
Was once used by servers supporting multiple regional encodings (e.g. ISO-8859-1 vs UTF-8) to choose which charset to encode a response in.
Example
Accept-Charset: utf-8, iso-8859-1;q=0.5History
Part of the original HTTP/1.1 specification, RFC 2068 (1997).
Deprecated / legacy status
Formally deprecated by RFC 7231 (2014) once UTF-8 became the de facto universal encoding for the web; modern browsers stopped sending it entirely, since assuming UTF-8 support almost never fails.