TEST #2633 RESULTS
02/14/2026 - 3:06 PM
ACCELERATOR
MODEL
48.0
tokens/s
generation
334
ms
time to first token
3965
tokens/s
prompt
829
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 3950X 16-Core Processor (znver2)
RAM
63.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
5046
tokens/s
47.8
tokens/s
225
ms
pp4096+tg256
3774
tokens/s
49.1
tokens/s
1.11
sec
pp2048+tg256
4581
tokens/s
48.2
tokens/s
468
ms
pp2048+tg768
4518
tokens/s
48.8
tokens/s
474
ms
pp1024+tg1024
5098
tokens/s
47.1
tokens/s
222
ms
pp1280+tg3072
4706
tokens/s
48.0
tokens/s
293
ms
pp384+tg1152
5241
tokens/s
46.3
tokens/s
121
ms
pp64+tg1024
2113
tokens/s
48.8
tokens/s
50
ms
pp16+tg1536
604
tokens/s
48.2
tokens/s
46
ms