TEST #1194 RESULTS
07/08/2025 - 6:58 PM
ACCELERATOR
MODEL
39.3
tokens/s
generation
812
ms
time to first token
1604
tokens/s
prompt
426
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen3 14B - Q3_K - Large
SYSTEM
CPU
AMD Ryzen Threadripper 3970X 32-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
1920
tokens/s
40.4
tokens/s
557
ms
pp4096+tg256
1506
tokens/s
35.9
tokens/s
2.75
sec
pp2048+tg256
1778
tokens/s
39.3
tokens/s
1.18
sec
pp2048+tg768
1790
tokens/s
38.9
tokens/s
1.17
sec
pp1024+tg1024
1908
tokens/s
39.5
tokens/s
562
ms
pp1280+tg3072
1813
tokens/s
38.3
tokens/s
733
ms
pp384+tg1152
1991
tokens/s
40.3
tokens/s
220
ms
pp64+tg1024
1379
tokens/s
40.6
tokens/s
70
ms
pp16+tg1536
349
tokens/s
40.1
tokens/s
69
ms