TEST #940 RESULTS
05/24/2025 - 2:14 AM
ACCELERATOR
MODEL
24.3
tokens/s
generation
9.09
sec
time to first token
809
tokens/s
prompt
129
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i7-13650HX (alderlake)
RAM
15.6GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
616
tokens/s
18.1
tokens/s
1.73
sec
pp4096+tg256
71
tokens/s
4.8
tokens/s
57.91
sec
pp2048+tg256
1024
tokens/s
33.4
tokens/s
2.03
sec
pp2048+tg768
757
tokens/s
21.6
tokens/s
2.73
sec
pp1024+tg1024
1685
tokens/s
31.6
tokens/s
639
ms
pp1280+tg3072
79
tokens/s
11.6
tokens/s
16.35
sec
pp384+tg1152
1578
tokens/s
33.3
tokens/s
273
ms
pp64+tg1024
1176
tokens/s
32.5
tokens/s
86
ms
pp16+tg1536
300
tokens/s
31.7
tokens/s
83
ms