TEST #1691 RESULTS
10/20/2025 - 10:04 PM
ACCELERATOR
MODEL
114
tokens/s
generation
66
ms
time to first token
19907
tokens/s
prompt
3252
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 9355P 32-Core Processor
RAM
1133.1GB
OS
Linux
Kernel Release
6.8.0-85-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #85-Ubuntu SMP PREEMPT_DYNAMIC Thu Sep 18 15:26:59 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
27018
tokens/s
123
tokens/s
46
ms
pp4096+tg256
22337
tokens/s
101
tokens/s
193
ms
pp2048+tg256
25434
tokens/s
113
tokens/s
89
ms
pp2048+tg768
25405
tokens/s
111
tokens/s
90
ms
pp1024+tg1024
27017
tokens/s
114
tokens/s
47
ms
pp1280+tg3072
23855
tokens/s
107
tokens/s
62
ms
pp384+tg1152
22319
tokens/s
118
tokens/s
26
ms
pp64+tg1024
4439
tokens/s
120
tokens/s
22
ms
pp16+tg1536
1339
tokens/s
120
tokens/s
19
ms