can-i-run-this-llm

Can I run this LLM?

Can a Mac Studio M3 Ultra (96GB) run Llama 4 Scout (MoE)?

Llama · See model card · estimated for a 96.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Llama 4 Scout (MoE) at Q3_K_M · 51.8 GB · 45–68 tok/s on a Mac Studio M3 Ultra (96GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K39.6 GB54–81 tok/sRuns great
Q3_K_M51.8 GB45–68 tok/sRuns great
Q4_K_M65.4 GB39–58 tok/sTight / slow
Q5_K_M76.6 GB34–51 tok/sWon't fit
Q6_K88.4 GB31–46 tok/sWon't fit
Q8_0114.5 GB25–38 tok/sWon't fit
FP16215.6 GB15–22 tok/sWon't fit

Ran Llama 4 Scout (MoE) on a Mac Studio M3 Ultra (96GB)? Add your real numbers →