gpu4.pics
data via LocalScore
NVIDIA GeForce RTX 4060 Ti
llamafile 0.9.1
16 GB VRAM NVIDIA
running Llama 3.2 1B Instruct· Q4_K_M· 1.5B
218
tokens / sec
2,328 · Excellent
Generation
218tok/s
higher is better
Prompt / prefill
9,337tok/s
higher is better
Time to first token
161 ms
lower is better
Efficiency
1.95 tok/s/W
Electricity cost
$0.024/Mtok
Power draw
112 W
All benchmarks on this rig
| Model | Quant | Gen | Prompt | TTFT | Power | Perf/W | $/Mtok | Score |
|---|---|---|---|---|---|---|---|---|
| Llama 3.2 1B Instruct1.5B | Q4_K_M | 218 | 9,337 | 161 ms | 112 W | 1.95 tok/s/W | $0.024/Mtok | 2,328 |