can-i-run-this-llm

Can I run this LLM?

Can a MacBook Pro M2 Max (96GB) run Kimi Linear 48B-A3B (MoE)?

Kimi · MIT · estimated for a 96.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Kimi Linear 48B-A3B (MoE) at Q8_0 · 52.2 GB · 56–84 tok/s on a MacBook Pro M2 Max (96GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K16.0 GB103–155 tok/sRuns great
Q3_K_M21.2 GB92–138 tok/sRuns great
Q4_K_M29.8 GB78–117 tok/sRuns great
Q5_K_M35.0 GB71–107 tok/sRuns great
Q6_K40.5 GB66–98 tok/sRuns great
Q8_052.2 GB56–84 tok/sRuns great
FP1698.2 GB35–53 tok/sWon't fit

Ran Kimi Linear 48B-A3B (MoE) on a MacBook Pro M2 Max (96GB)? Add your real numbers →