TEST #3118 RESULTS
03/19/2026 - 12:31 PM
ACCELERATOR
MODEL
39.2
tokens/s
generation
1.49
sec
time to first token
975
tokens/s
prompt
295
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 3900X 12-Core Processor (znver2)
RAM
31.9GB
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
1168
tokens/s
46.1
tokens/s
899
ms
pp4096+tg256
728
tokens/s
22.9
tokens/s
5.67
sec
pp2048+tg256
974
tokens/s
34.1
tokens/s
2.13
sec
pp2048+tg768
975
tokens/s
31.9
tokens/s
2.13
sec
pp1024+tg1024
1152
tokens/s
39.3
tokens/s
912
ms
pp1280+tg3072
1112
tokens/s
28.9
tokens/s
1.18
sec
pp384+tg1152
1287
tokens/s
46.3
tokens/s
317
ms
pp64+tg1024
865
tokens/s
53.2
tokens/s
90
ms
pp16+tg1536
514
tokens/s
50.1
tokens/s
47
ms