TEST #2049 RESULTS
12/27/2025 - 8:31 AM
ACCELERATOR
MODEL
15.8
tokens/s
generation
1.26
sec
time to first token
1018
tokens/s
prompt
234
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
RAM
119.7GB
OS
Linux
Kernel Release
6.14.0-1015-nvidia
Architecture
aarch64
Version
Cosmopolitan 3.9.7 MODE=aarch64; #15-Ubuntu SMP PREEMPT_DYNAMIC Tue Nov 25 18:02:16 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1386
tokens/s
18.3
tokens/s
794
ms
pp4096+tg256
968
tokens/s
16.2
tokens/s
4.29
sec
pp2048+tg256
1223
tokens/s
17.4
tokens/s
1.73
sec
pp2048+tg768
1220
tokens/s
17.1
tokens/s
1.74
sec
pp1024+tg1024
1242
tokens/s
17.0
tokens/s
880
ms
pp1280+tg3072
1289
tokens/s
16.1
tokens/s
1.07
sec
pp384+tg1152
1098
tokens/s
15.0
tokens/s
424
ms
pp64+tg1024
582
tokens/s
12.7
tokens/s
216
ms
pp16+tg1536
151
tokens/s
12.4
tokens/s
181
ms