Can I run this LLM?
Can a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB) run Qwen3.8 Flash Next (MoE)?
Qwen · See model card · estimated for a 144.0 GB machine at 4K context.
Yes — it runs great.
Best fit: Qwen3.8 Flash Next (MoE) at FP16 · 354.0 GB · 32–49 tok/s on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB).
Every quantization on this machine
| Quant | Size | Speed | Fit |
|---|---|---|---|
| Q2_K | 80.9 GB | 285–427 tok/s | Runs great |
| Q3_K_M | 92.0 GB | 276–415 tok/s | Runs great |
| Q4_K_M | 119.6 GB | 258–387 tok/s | Runs great |
| Q5_K_M | 134.7 GB | 214–321 tok/s | Runs great |
| Q6_K | 168.1 GB | 115–173 tok/s | Runs great |
| Q8_0 | 188.3 GB | 90–135 tok/s | Runs great |
| FP16 | 354.0 GB | 32–49 tok/s | Runs great |
Open the full calculator →Everything a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB) runsQwen3.8 Flash Next (MoE) on other machinesWhat to buy to run it →
Ran Qwen3.8 Flash Next (MoE) on a NVIDIA GH200 Grace Hopper (144GB HBM + 480GB)? Add your real numbers →