TEST #3675 RESULTS
06/03/2026 - 1:05 PM
ACCELERATOR
MODEL
7.9
tokens/s
generation
43.74
sec
time to first token
32
tokens/s
prompt
18
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
12th Gen Intel Core i5-12400 (alderlake)
RAM
31GB
OS
Linux
Kernel Release
7.0.3-arch1-2
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Fri, 01 May 2026 15:49:22 +0000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
33
tokens/s
8.3
tokens/s
31.39
sec
pp4096+tg256
29
tokens/s
6.9
tokens/s
142.43
sec
pp2048+tg256
29
tokens/s
7.8
tokens/s
69.73
sec
pp2048+tg768
31
tokens/s
7.7
tokens/s
65.32
sec
pp1024+tg1024
33
tokens/s
8.0
tokens/s
31.33
sec
pp1280+tg3072
32
tokens/s
7.5
tokens/s
39.59
sec
pp384+tg1152
34
tokens/s
8.2
tokens/s
11.31
sec
pp64+tg1024
34
tokens/s
8.4
tokens/s
1.98
sec
pp16+tg1536
34
tokens/s
8.3
tokens/s
586
ms