TEST #2024 RESULTS
12/22/2025 - 8:13 AM
ACCELERATOR
MODEL
73.9
tokens/s
generation
1.30
sec
time to first token
992
tokens/s
prompt
384
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 4P+6E
RAM
24GB
OS
Darwin
Kernel Release
25.0.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.0.0: Wed Sep 17 21:42:08 PDT 2025; root:xnu-12377.1.9~141/RELEASE_ARM64_T8132
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1204
tokens/s
82.1
tokens/s
863
ms
pp4096+tg256
906
tokens/s
53.2
tokens/s
4.54
sec
pp2048+tg256
1095
tokens/s
68.7
tokens/s
1.88
sec
pp2048+tg768
1091
tokens/s
66.3
tokens/s
1.89
sec
pp1024+tg1024
1178
tokens/s
75.5
tokens/s
881
ms
pp1280+tg3072
1143
tokens/s
62.4
tokens/s
1.13
sec
pp384+tg1152
1187
tokens/s
83.9
tokens/s
334
ms
pp64+tg1024
900
tokens/s
88.3
tokens/s
81
ms
pp16+tg1536
222
tokens/s
85.3
tokens/s
82
ms