can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB) run Qwen3 Coder Next 80B-A3B (MoE)?

Qwen · Apache 2.0 · estimated for a 144.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Qwen3 Coder Next 80B-A3B (MoE) at FP16 · 159.5 GB · 121–181 tok/s on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K29.2 GB318–478 tok/sRuns great
Q3_K_M38.3 GB309–464 tok/sRuns great
Q4_K_M49.3 GB299–448 tok/sRuns great
Q5_K_M59.2 GB290–434 tok/sRuns great
Q6_K65.8 GB284–426 tok/sRuns great
Q8_084.8 GB269–403 tok/sRuns great
FP16159.5 GB121–181 tok/sRuns great

Ran Qwen3 Coder Next 80B-A3B (MoE) on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB)? Add your real numbers →