can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA A100 (80GB) run Llama 4 Scout (MoE)?

Llama · See model card · estimated for a 80.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Llama 4 Scout (MoE) at Q3_K_M · 51.8 GB · 115–173 tok/s on a NVIDIA A100 (80GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K39.6 GB135–203 tok/sRuns great
Q3_K_M51.8 GB115–173 tok/sRuns great
Q4_K_M65.4 GB99–148 tok/sTight / slow
Q5_K_M76.6 GB89–133 tok/sWon't fit
Q6_K88.4 GB80–120 tok/sWon't fit
Q8_0114.5 GB66–98 tok/sWon't fit
FP16215.6 GB39–58 tok/sWon't fit

Ran Llama 4 Scout (MoE) on a NVIDIA A100 (80GB)? Add your real numbers →