TEST #1758 RESULTS
11/08/2025 - 6:56 AM
ACCELERATOR
MODEL
9.2
tokens/s
generation
14.52
sec
time to first token
90
tokens/s
prompt
39
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 4P+6E
RAM
16GB
OS
Darwin
Kernel Release
25.1.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.1.0: Mon Oct 20 19:32:56 PDT 2025; root:xnu-12377.41.6~2/RELEASE_ARM64_T8132
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
104
tokens/s
9.5
tokens/s
9.96
sec
pp4096+tg256
88
tokens/s
7.5
tokens/s
46.94
sec
pp2048+tg256
91
tokens/s
8.8
tokens/s
22.74
sec
pp2048+tg768
93
tokens/s
9.0
tokens/s
22.02
sec
pp1024+tg1024
100
tokens/s
9.6
tokens/s
10.36
sec
pp1280+tg3072
95
tokens/s
8.7
tokens/s
13.63
sec
pp384+tg1152
102
tokens/s
9.9
tokens/s
3.86
sec
pp64+tg1024
97
tokens/s
9.9
tokens/s
758
ms
pp16+tg1536
45
tokens/s
9.9
tokens/s
448
ms