TEST #3649 RESULTS
05/29/2026 - 11:49 AM
ACCELERATOR
MODEL
13.6
tokens/s
generation
17.61
sec
time to first token
80
tokens/s
prompt
39
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i9-14900K (alderlake)
RAM
62.1GB
OS
Linux
Kernel Release
6.12.90+deb13-amd64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Debian 6.12.90-1 (2026-05-22)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
82
tokens/s
14.0
tokens/s
12.62
sec
pp4096+tg256
71
tokens/s
12.5
tokens/s
57.45
sec
pp2048+tg256
77
tokens/s
13.4
tokens/s
26.75
sec
pp2048+tg768
76
tokens/s
13.3
tokens/s
26.86
sec
pp1024+tg1024
80
tokens/s
13.7
tokens/s
12.80
sec
pp1280+tg3072
79
tokens/s
13.0
tokens/s
16.22
sec
pp384+tg1152
83
tokens/s
13.9
tokens/s
4.70
sec
pp64+tg1024
84
tokens/s
14.2
tokens/s
829
ms
pp16+tg1536
83
tokens/s
14.0
tokens/s
263
ms