UTF-8 Solr answer yields charlist

Currently I have a problem with an answer I’m getting from my Solr server.

When I request it directly from JavaScript or with Postman/Insomnia I get an UTF-8 document that displays my umlauts correctly. E. g. one value contains the german word Tür which will be rendered as is.

When I make the request from Elixir I’ll get a charlist as the documents body. This is then parsed as JSON and theoretically working as expected, except that now the Tür gets rendered as Tür.

What am I doing wrong here? I already tried converting it to string, which only yields a bitstring.