TEST #1678 RESULTS
10/18/2025 - 6:59 PM
ACCELERATOR
MODEL
20.6
tokens/s
generation
3.01
sec
time to first token
427
tokens/s
prompt
143
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen Threadripper 1920X 12-Core Processor
RAM
62.6GB
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
478
tokens/s
20.5
tokens/s
2.19
sec
pp4096+tg256
424
tokens/s
16.0
tokens/s
9.72
sec
pp2048+tg256
461
tokens/s
19.3
tokens/s
4.49
sec
pp2048+tg768
454
tokens/s
19.2
tokens/s
4.56
sec
pp1024+tg1024
481
tokens/s
21.3
tokens/s
2.17
sec
pp1280+tg3072
468
tokens/s
18.5
tokens/s
2.78
sec
pp384+tg1152
489
tokens/s
22.8
tokens/s
828
ms
pp64+tg1024
390
tokens/s
24.0
tokens/s
201
ms
pp16+tg1536
200
tokens/s
23.4
tokens/s
119
ms