Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
314.8 tok/s
TTFT
615 ms
Safe context
63K
Memory
46.0 GB / 180.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 314.8 tok/s | 895 ms | 112K |
| Chat | C | Runs well | 314.8 tok/s | 350 ms | 33K |
| Coding | C | Runs well | 314.8 tok/s | 615 ms | 63K |
| RAG | C | Runs well | 314.8 tok/s | 1118 ms | 112K |
| Reasoning | C | Runs well | 314.8 tok/s | 727 ms | 63K |
How Qwen3.5 35B A3B (35B params) fits at each quantization level on NVIDIA B200 180GB (180.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 13.7 GB | Low | D31 |
Q3_K_S | 3 | 17.2 GB | Low | D31 |
NVFP4 | 4 | 19.6 GB | Medium | D32 |
Q4_K_M | 4 | 21.3 GB | Medium | D32 |
Q5_K_M | 5 | 25.2 GB | High | D32 |
Q6_K | 6 | 28.7 GB | High | D33 |
Q8_0 | 8 | 37.5 GB | Very High | D34 |
F16Best for your GPU | 16 | 71.8 GB | Maximum | D38 |
huggingface-cli download hf-lmstudio-community--qwen3-5-35b-a3b-gguf