TEST #1542 RESULTS
09/22/2025 - 6:06 PM
ACCELERATOR
MODEL
26.2
tokens/s
generation
546
ms
time to first token
2433
tokens/s
prompt
489
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i7-13700HX (alderlake)
RAM
63.7GB
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
2979
tokens/s
25.3
tokens/s
385
ms
pp4096+tg256
2332
tokens/s
26.9
tokens/s
1.78
sec
pp2048+tg256
2748
tokens/s
25.9
tokens/s
789
ms
pp2048+tg768
2716
tokens/s
26.0
tokens/s
791
ms
pp1024+tg1024
3033
tokens/s
26.1
tokens/s
378
ms
pp1280+tg3072
2908
tokens/s
26.0
tokens/s
479
ms
pp384+tg1152
2940
tokens/s
26.5
tokens/s
162
ms
pp64+tg1024
1887
tokens/s
25.5
tokens/s
74
ms
pp16+tg1536
351
tokens/s
27.0
tokens/s
72
ms