TEST #3395 RESULTS
04/26/2026 - 8:46 PM
ACCELERATOR
MODEL
33.2
tokens/s
generation
1.23
sec
time to first token
1062
tokens/s
prompt
306
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7251 8-Core Processor
RAM
220.3GB
OS
Linux
Kernel Release
6.17.0-22-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #22~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Mar 26 15:25:54 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1258
tokens/s
34.9
tokens/s
843
ms
pp4096+tg256
1020
tokens/s
30.6
tokens/s
4.05
sec
pp2048+tg256
1156
tokens/s
33.1
tokens/s
1.80
sec
pp2048+tg768
1143
tokens/s
32.5
tokens/s
1.82
sec
pp1024+tg1024
1210
tokens/s
33.5
tokens/s
876
ms
pp1280+tg3072
1151
tokens/s
31.5
tokens/s
1.14
sec
pp384+tg1152
1248
tokens/s
34.1
tokens/s
336
ms
pp64+tg1024
1053
tokens/s
34.7
tokens/s
116
ms
pp16+tg1536
323
tokens/s
34.3
tokens/s
86
ms