TEST #956 RESULTS
05/30/2025 - 7:49 PM
ACCELERATOR
MODEL
8.1
tokens/s
generation
16.85
sec
time to first token
71
tokens/s
prompt
32
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 4P+6E
RAM
32GB
OS
Darwin
Kernel Release
24.5.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.5.0: Tue Apr 22 19:54:43 PDT 2025; root:xnu-11417.121.6~2/RELEASE_ARM64_T8132
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
94
tokens/s
10.0
tokens/s
10.96
sec
pp4096+tg256
82
tokens/s
7.8
tokens/s
50.28
sec
pp2048+tg256
84
tokens/s
8.6
tokens/s
24.48
sec
pp2048+tg768
77
tokens/s
7.7
tokens/s
26.62
sec
pp1024+tg1024
74
tokens/s
7.9
tokens/s
14.02
sec
pp1280+tg3072
72
tokens/s
7.0
tokens/s
17.91
sec
pp384+tg1152
72
tokens/s
7.9
tokens/s
5.45
sec
pp64+tg1024
54
tokens/s
8.2
tokens/s
1.31
sec
pp16+tg1536
32
tokens/s
8.1
tokens/s
604
ms
