gpu4.pics
data via LocalScore
NVIDIA Tesla V100
llamafile 0.9.2
32 GB VRAM NVIDIA 300 W TDP
running Llama 3.2 1B Instruct· Q4_K_M· 1.5B
248
tokens / sec
2,545 · Excellent
Generation
248tok/s
higher is better
Prompt / prefill
8,984tok/s
higher is better
Time to first token
135 ms
lower is better
Composite score
2,545
Model size
1.5B
Runtime
llamafile
All benchmarks on this rig
| Model | Quant | Gen | Prompt | TTFT | Power | Perf/W | $/Mtok | Score |
|---|---|---|---|---|---|---|---|---|
| Llama 3.2 1B Instruct1.5B | Q4_K_M | 248 | 8,984 | 135 ms | — | — | — | 2,545 |