TEST #482 RESULTS
04/07/2025 - 9:34 AM
ACCELERATOR
MODEL
68.0
tokens/s
generation
370
ms
time to first token
3682
tokens/s
prompt
878
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
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4482
tokens/s
69.8
tokens/s
243
ms
pp4096+tg256
3151
tokens/s
63.9
tokens/s
1.32
sec
pp2048+tg256
3963
tokens/s
67.3
tokens/s
531
ms
pp2048+tg768
3983
tokens/s
67.7
tokens/s
529
ms
pp1024+tg1024
4534
tokens/s
68.7
tokens/s
240
ms
pp1280+tg3072
4405
tokens/s
66.0
tokens/s
305
ms
pp384+tg1152
5005
tokens/s
69.1
tokens/s
91
ms
pp64+tg1024
2822
tokens/s
69.7
tokens/s
37
ms
pp16+tg1536
797
tokens/s
69.4
tokens/s
34
ms