TEST #1439 RESULTS
08/28/2025 - 1:33 PM
ACCELERATOR
MODEL
50.6
tokens/s
generation
521
ms
time to first token
2599
tokens/s
prompt
632
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 7 265 (arrowlake-s)
RAM
31.5GB
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
3170
tokens/s
50.8
tokens/s
344
ms
pp4096+tg256
2275
tokens/s
49.0
tokens/s
1.82
sec
pp2048+tg256
2791
tokens/s
50.7
tokens/s
755
ms
pp2048+tg768
2792
tokens/s
50.6
tokens/s
754
ms
pp1024+tg1024
3109
tokens/s
51.2
tokens/s
349
ms
pp1280+tg3072
3017
tokens/s
49.2
tokens/s
446
ms
pp384+tg1152
3494
tokens/s
51.1
tokens/s
129
ms
pp64+tg1024
2092
tokens/s
51.3
tokens/s
50
ms
pp16+tg1536
652
tokens/s
51.2
tokens/s
44
ms