gpu4.pics
data via LocalScore
NVIDIA RTX 3090 Ti
24 GB VRAM NVIDIA 450 W TDP
running Qwen2.5 14B Instruct· Q4_K_M· 14.8B
64.2
tokens / sec
627 · Strong #12 of 373
Generation
64.2tok/s
higher is better
Prompt / prefill
2,202tok/s
higher is better
Time to first token
575 ms
lower is better
Composite score
627
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 | 64.2 | 2,202 | 575 ms | — | — | — | 627 |