TEST #2025 RESULTS
12/22/2025 - 8:20 AM
ACCELERATOR
MODEL
74.6
tokens/s
generation
1.29
sec
time to first token
1009
tokens/s
prompt
387
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 4P+6E
RAM
16GB
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
1201
tokens/s
83.1
tokens/s
865
ms
pp4096+tg256
910
tokens/s
53.3
tokens/s
4.52
sec
pp2048+tg256
1094
tokens/s
69.0
tokens/s
1.89
sec
pp2048+tg768
1080
tokens/s
66.5
tokens/s
1.91
sec
pp1024+tg1024
1185
tokens/s
76.2
tokens/s
876
ms
pp1280+tg3072
1166
tokens/s
63.0
tokens/s
1.11
sec
pp384+tg1152
1197
tokens/s
83.9
tokens/s
331
ms
pp64+tg1024
899
tokens/s
89.9
tokens/s
82
ms
pp16+tg1536
349
tokens/s
86.3
tokens/s
56
ms