TEST #1825 RESULTS
11/19/2025 - 12:15 PM
ACCELERATOR
MODEL
46.0
tokens/s
generation
686
ms
time to first token
1885
tokens/s
prompt
502
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon Gold 5220R CPU @ 2.20GHz (cascadelake)
RAM
63.6GB
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
2294
tokens/s
48.5
tokens/s
467
ms
pp4096+tg256
1822
tokens/s
44.9
tokens/s
2.27
sec
pp2048+tg256
2096
tokens/s
46.0
tokens/s
1000
ms
pp2048+tg768
2084
tokens/s
45.7
tokens/s
1.01
sec
pp1024+tg1024
2217
tokens/s
45.7
tokens/s
483
ms
pp1280+tg3072
2111
tokens/s
45.1
tokens/s
630
ms
pp384+tg1152
2324
tokens/s
45.9
tokens/s
188
ms
pp64+tg1024
1661
tokens/s
46.3
tokens/s
60
ms
pp16+tg1536
358
tokens/s
46.0
tokens/s
67
ms