gpu4.pics
data via LocalScore
NVIDIA H100
80 GB VRAM NVIDIA 700 W TDP
running Qwen2.5 14B Instruct· Q4_K_M· 14.8B
72.2
tokens / sec
802 · Strong #8 of 373
Generation
72.2tok/s
higher is better
Prompt / prefill
2,973tok/s
higher is better
Time to first token
417 ms
lower is better
Composite score
802
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 | 72.2 | 2,973 | 417 ms | — | — | — | 802 |