TEST #3314 RESULTS
04/15/2026 - 1:38 PM
ACCELERATOR
MODEL
110
tokens/s
generation
672
ms
time to first token
1806
tokens/s
prompt
667
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 Pro 8P+4E
RAM
48GB
OS
Darwin
Kernel Release
25.4.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.4.0: Thu Mar 19 19:33:25 PDT 2026; root:xnu-12377.101.15~1/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2224
tokens/s
124
tokens/s
468
ms
pp4096+tg256
1865
tokens/s
74.0
tokens/s
2.21
sec
pp2048+tg256
2087
tokens/s
100
tokens/s
991
ms
pp2048+tg768
2090
tokens/s
95.9
tokens/s
990
ms
pp1024+tg1024
2167
tokens/s
113
tokens/s
481
ms
pp1280+tg3072
2147
tokens/s
90.3
tokens/s
605
ms
pp384+tg1152
2079
tokens/s
127
tokens/s
192
ms
pp64+tg1024
1234
tokens/s
138
tokens/s
58
ms
pp16+tg1536
360
tokens/s
131
tokens/s
51
ms