Seeing some good stuff in the comments, but how are people doing this?
I am not, but I can tell you how:
homelab running liteLLM or gomodel