Skip to content
mlx.app

Compare

Two things worth comparing: which model to run on the Mac you have, and which Mac to buy for the models you want.

8.0B · bf16FITS · 21.3 GB FREE
Weights
16.1 GB
KV cache
1.1 GB
Speed est.
13.3 tok/s
Max context
128K
Licence
llama-3.1
mlx_lm.chat --model mlx-community/Meta-Llama-3.1-8B-Instruct-bf16
14.8B · q8FITS · 21.3 GB FREE
Weights
15.7 GB
KV cache
1.3 GB
Speed est.
13.5 tok/s
Max context
40K
Licence
apache-2.0
mlx_lm.chat --model mlx-community/Qwen3-14B-8bit