gpu4.pics
data via LocalScore
NVIDIA RTX PRO 6000 Blackwell Server Edition
95 GB VRAM NVIDIA
running Qwen2.5 14B Instruct· Q4_K_M· 14.8B
17.9
tokens / sec
438 · Passable #32 of 373
Generation
17.9tok/s
higher is better
Prompt / prefill
2,581tok/s
higher is better
Time to first token
637 ms
lower is better
Composite score
438
Model size
14.8B
Runtime
—
All benchmarks on this rig
| Model | Quant | Gen | Prompt | TTFT | Power | Perf/W | $/Mtok | Score |
|---|---|---|---|---|---|---|---|---|
| Qwen2.5 14B Instruct14.8B | Q4_K_M | 17.9 | 2,581 | 637 ms | — | — | — | 438 |