[pzo]

Ollama is a wrapper around llama.cpp thei using ggml format. Onnx is different ml model format and onnxruntime developer by microsoft. Mlx is ml framework from Apple. If you want the fastest speed on MacOS most likely stick with mlx