TEST #4278 RESULTS
10/13/2026 - 7:02 PM
ACCELERATOR
MODEL
6.4
tokens/s
generation
14.93
sec
time to first token
92
tokens/s
prompt
34
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon Platinum 8269CY CPU @ 2.50GHz (cascadelake)
RAM
93.9GB
OS
Linux
Kernel Release
7.0.0-38-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #38~24.04.4-Ubuntu SMP PREEMPT_DYNAMIC Mon Sep 14 16:37:11 UTC 2
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
96
tokens/s
6.6
tokens/s
10.79
sec
pp4096+tg256
83
tokens/s
5.9
tokens/s
49.30
sec
pp2048+tg256
90
tokens/s
6.4
tokens/s
22.86
sec
pp2048+tg768
92
tokens/s
6.3
tokens/s
22.46
sec
pp1024+tg1024
97
tokens/s
6.5
tokens/s
10.66
sec
pp1280+tg3072
98
tokens/s
6.2
tokens/s
13.28
sec
pp384+tg1152
103
tokens/s
6.6
tokens/s
3.87
sec
pp64+tg1024
97
tokens/s
6.7
tokens/s
809
ms
pp16+tg1536
69
tokens/s
6.6
tokens/s
377
ms