MLX (mlx-lm)
Apple's own framework that parks big models on unified memory via 4-bit MLX quants — the pragmatic choice when your only GPU is a Mac.
https://github.com/ml-explore/mlx-lm
metalsafetensorsMLX
No benchmarks pasted on MLX (mlx-lm) yet. Paste one — it takes one keystroke and no form.