NVIDIA B200 180GBBudget pickB88.8 tok/s decode
~$30,000 MSRP
Can it run?
Runs well
Using Q4_K_M in vLLM
Fit status
Runs well
Decode
45.7 tok/s
TTFT
4237 ms
Safe context
20K
Memory
111.5 GB / 141.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Tight fit | 53.3 tok/s | 5283 ms | 34K |
| Chat | B | Runs well | 53.3 tok/s | 1981 ms | 11K |
| Coding | C | Runs well | 45.7 tok/s | 4237 ms | 20K |
| RAG | C | Tight fit | 53.3 tok/s | 6603 ms | 34K |
| Reasoning | B | Runs well | 53.3 tok/s | 4292 ms | 20K |
How Pixtral Large 124B (124B params) fits at each quantization level on NVIDIA H200 141GB (141.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 48.4 GB | Low | D37 |
Q3_K_S | 3 | 60.8 GB | Low | D39 |
NVFP4 | 4 |
huggingface-cli download pixtral-large-124bUpgrade options
NVIDIA B200 180GBBudget pick~$30,000 MSRP
69.4 GB |
| Medium |
| D40 |
Q4_K_M | 4 | 75.6 GB | Medium | C41 |
Q5_K_M | 5 | 89.3 GB | High | C43 |
Q6_KBest for your GPU | 6 | 101.7 GB | High | C44 |
Q8_0 | 8 | 132.7 GB | Very High | C44 |
F16 | 16 | 254.2 GB | Maximum | F0 |