Why are you using Ollama? Just use llama.cpp
brew install llama.cpp
use the inbuilt CLI, Server or Chat interface. + Hook it up to any other app
For MLX I'd guess.
For MLX I'd guess.