can-i-run-this-llm

Can I run this LLM?

Can a NVIDIA B200 (192GB) run Hunyuan Hy3 (MoE)?

Hunyuan · Apache 2.0 · estimated for a 192.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Hunyuan Hy3 (MoE) at Q3_K_M · 128.9 GB · 223–334 tok/s on a NVIDIA B200 (192GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K97.1 GB243–365 tok/sRuns great
Q3_K_M128.9 GB223–334 tok/sRuns great
Q4_K_M181.1 GB196–294 tok/sWon't fit
Q5_K_M212.9 GB182–273 tok/sWon't fit
Q6_K246.5 GB170–255 tok/sWon't fit
Q8_0317.5 GB149–223 tok/sWon't fit

Ran Hunyuan Hy3 (MoE) on a NVIDIA B200 (192GB)? Add your real numbers →