Thanks! I assumed simpler == faster, but my ignorance is showing itself.

I am using the model they recommended in the blog post - which I assumed was using MLX?