TEST #1925 RESULTS
12/10/2025 - 1:18 PM
ACCELERATOR
MODEL
65.3
tokens/s
generation
689
ms
time to first token
1870
tokens/s
prompt
562
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
13th Gen Intel Core i5-13500 (alderlake)
RAM
31.8GB
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
2210
tokens/s
67.6
tokens/s
478
ms
pp4096+tg256
1800
tokens/s
60.0
tokens/s
2.29
sec
pp2048+tg256
2048
tokens/s
65.2
tokens/s
1.02
sec
pp2048+tg768
2031
tokens/s
64.6
tokens/s
1.02
sec
pp1024+tg1024
2196
tokens/s
66.2
tokens/s
481
ms
pp1280+tg3072
2103
tokens/s
62.5
tokens/s
623
ms
pp384+tg1152
2273
tokens/s
67.0
tokens/s
185
ms
pp64+tg1024
1627
tokens/s
68.0
tokens/s
54
ms
pp16+tg1536
544
tokens/s
66.9
tokens/s
45
ms