can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA H100 (80GB) run Qwen3.8 27B?

Qwen · Apache 2.0 · estimated for a 80.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Qwen3.8 27B at FP16 · 54.6 GB · 40–59 tok/s on a NVIDIA H100 (80GB).

Every quantization on this machine

QuantSizeSpeedFit
Q3_K_M11.8 GB126–189 tok/sRuns great
Q4_K_M16.6 GB101–152 tok/sRuns great
Q5_K_M19.5 GB91–136 tok/sRuns great
Q6_K22.5 GB81–122 tok/sRuns great
Q8_029.0 GB67–101 tok/sRuns great
FP1654.6 GB40–59 tok/sRuns great

Ran Qwen3.8 27B on a NVIDIA H100 (80GB)? Add your real numbers →