TEST #1773 RESULTS
11/09/2025 - 11:19 PM
ACCELERATOR
MODEL
18.8
tokens/s
generation
1.19
sec
time to first token
1152
tokens/s
prompt
264
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
RAM
119.7GB
OS
Linux
Kernel Release
6.11.0-1016-nvidia
Architecture
aarch64
Version
Cosmopolitan 3.9.7 MODE=aarch64; #16-Ubuntu SMP PREEMPT_DYNAMIC Sun Sep 21 16:52:46 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1422
tokens/s
19.2
tokens/s
775
ms
pp4096+tg256
1000
tokens/s
17.3
tokens/s
4.16
sec
pp2048+tg256
1248
tokens/s
19.0
tokens/s
1.70
sec
pp2048+tg768
1245
tokens/s
18.4
tokens/s
1.70
sec
pp1024+tg1024
1404
tokens/s
19.0
tokens/s
785
ms
pp1280+tg3072
1354
tokens/s
17.9
tokens/s
1.00
sec
pp384+tg1152
1567
tokens/s
19.4
tokens/s
300
ms
pp64+tg1024
867
tokens/s
19.8
tokens/s
141
ms
pp16+tg1536
265
tokens/s
19.6
tokens/s
112
ms