can-i-run-this-llm

Can I run this LLM?

Can a Mac mini M6 (32GB) run Kimi Linear 48B-A3B (MoE)?

Kimi · MIT · estimated for a 32.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Kimi Linear 48B-A3B (MoE) at Q2_K · 17.5 GB · 65–97 tok/s on a Mac mini M6 (32GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K17.5 GB65–97 tok/sRuns great
Q3_K_M22.7 GB55–82 tok/sWon't fit
Q4_K_M30.1 GB45–68 tok/sWon't fit
Q5_K_M35.1 GB40–60 tok/sWon't fit
Q6_K40.5 GB36–54 tok/sWon't fit
Q8_052.3 GB29–44 tok/sWon't fit
FP1698.3 GB17–26 tok/sWon't fit

Ran Kimi Linear 48B-A3B (MoE) on a Mac mini M6 (32GB)? Add your real numbers →