Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
201.2 tok/s
TTFT
962 ms
Safe context
25K
Memory
60.4 GB / 96.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | B | Runs well | 201.2 tok/s | 1400 ms | 51K |
| Chat | B | Runs well | 201.2 tok/s | 525 ms | 13K |
| Coding | B | Runs well | 201.2 tok/s | 962 ms | 25K |
| RAG | B | Runs well | 201.2 tok/s | 1750 ms | 51K |
| Reasoning | B | Runs well | 201.2 tok/s | 1137 ms | 25K |
How Qwen3-Coder-Next (80B params) fits at each quantization level on NVIDIA GH200 96GB (96.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 31.2 GB | Low | D37 |
Q3_K_S | 3 | 39.2 GB | Low | D38 |
NVFP4 | 4 | 44.8 GB | Medium | D40 |
Q4_K_M | 4 | 48.8 GB | Medium | C40 |
Q5_K_M | 5 | 57.6 GB | High | C42 |
Q6_KBest for your GPU | 6 | 65.6 GB | High | C44 |
Q8_0 | 8 | 85.6 GB | Very High | C45 |
F16 | 16 | 164.0 GB | Maximum | F0 |
ollama run qwen-3-coder-nexthuggingface-cli download qwen-3-coder-next