TEST #3287 RESULTS
04/11/2026 - 9:38 AM
ACCELERATOR
MODEL
98.4
tokens/s
generation
182
ms
time to first token
7159
tokens/s
prompt
1570
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i5-13600K (alderlake)
RAM
31.8GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
9430
tokens/s
98.6
tokens/s
119
ms
pp4096+tg256
6629
tokens/s
93.7
tokens/s
628
ms
pp2048+tg256
8260
tokens/s
96.4
tokens/s
258
ms
pp2048+tg768
8267
tokens/s
97.6
tokens/s
258
ms
pp1024+tg1024
9456
tokens/s
97.9
tokens/s
118
ms
pp1280+tg3072
9121
tokens/s
96.3
tokens/s
151
ms
pp384+tg1152
8050
tokens/s
98.7
tokens/s
59
ms
pp64+tg1024
4039
tokens/s
103
tokens/s
26
ms
pp16+tg1536
1179
tokens/s
103
tokens/s
24
ms