The stability of the language is such a blessing.
replyI think that's part of the reason that LLMs do so well with it, despite its relative lack of popularity.
deleted
replyWhich models do you use for elixir?
replyAll the leading ones I've tried work mostly fine: Claude Sonnet 4.5, Opus 4.5/6/8, GPT 5.4/5, Kimi K2.5/6, GLM 5.1, and so on.
replyThey can all write serviceable Elixir. Opus is my preferred one, but they do decently well enough for typical coding tasks.
Claude seems to work well.
reply