3 ms·

Will this perform better than lmstudio-community/Qwen3.8-27B-MLX-4bit on my m5 max 48gb memory mac?