TEST #3737 RESULTS
06/16/2026 - 4:02 PM
ACCELERATOR
MODEL
46.4
tokens/s
generation
557
ms
time to first token
2349
tokens/s
prompt
580
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 7 265H (arrowlake)
RAM
31.4GB
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
2855
tokens/s
46.7
tokens/s
380
ms
pp4096+tg256
2201
tokens/s
46.0
tokens/s
1.88
sec
pp2048+tg256
2591
tokens/s
44.4
tokens/s
814
ms
pp2048+tg768
2591
tokens/s
44.3
tokens/s
814
ms
pp1024+tg1024
2850
tokens/s
46.8
tokens/s
381
ms
pp1280+tg3072
2742
tokens/s
45.5
tokens/s
489
ms
pp384+tg1152
3145
tokens/s
47.7
tokens/s
143
ms
pp64+tg1024
1624
tokens/s
47.7
tokens/s
60
ms
pp16+tg1536
539
tokens/s
48.3
tokens/s
50
ms