TEST #4097 RESULTS
08/28/2026 - 5:34 PM
ACCELERATOR
MODEL
23.3
tokens/s
generation
9.44
sec
time to first token
146
tokens/s
prompt
71
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Genuine Intel CPU 0000%@ (sapphirerapids)
RAM
125.5GB
OS
Linux
Kernel Release
6.12.105+deb13-amd64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Debian 6.12.105-1 (2026-08-24)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
148
tokens/s
25.8
tokens/s
6.98
sec
pp4096+tg256
130
tokens/s
16.7
tokens/s
31.58
sec
pp2048+tg256
147
tokens/s
22.0
tokens/s
13.99
sec
pp2048+tg768
147
tokens/s
26.8
tokens/s
13.93
sec
pp1024+tg1024
150
tokens/s
21.4
tokens/s
6.87
sec
pp1280+tg3072
150
tokens/s
26.1
tokens/s
8.57
sec
pp384+tg1152
164
tokens/s
19.3
tokens/s
2.38
sec
pp64+tg1024
154
tokens/s
24.6
tokens/s
451
ms
pp16+tg1536
126
tokens/s
26.8
tokens/s
166
ms