Ah sure that's my bad then. It just runs incredibly poorly on my end, so I guess my default was to assume it was an LLM issue.