TEST #3511 RESULTS
05/11/2026 - 10:11 PM
ACCELERATOR
MODEL
58.8
tokens/s
generation
679
ms
time to first token
1932
tokens/s
prompt
551
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7251 8-Core Processor
RAM
220.3GB
OS
Linux
Kernel Release
6.17.0-23-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #23~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Apr 14 16:11:48 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2271
tokens/s
62.2
tokens/s
467
ms
pp4096+tg256
1824
tokens/s
54.1
tokens/s
2.26
sec
pp2048+tg256
2088
tokens/s
58.6
tokens/s
998
ms
pp2048+tg768
2083
tokens/s
57.6
tokens/s
1.00
sec
pp1024+tg1024
2215
tokens/s
59.4
tokens/s
479
ms
pp1280+tg3072
2115
tokens/s
55.6
tokens/s
622
ms
pp384+tg1152
2316
tokens/s
60.0
tokens/s
182
ms
pp64+tg1024
1989
tokens/s
61.0
tokens/s
48
ms
pp16+tg1536
492
tokens/s
60.4
tokens/s
48
ms