TEST #3238 RESULTS
04/05/2026 - 9:24 PM
ACCELERATOR
MODEL
66.8
tokens/s
generation
1.41
sec
time to first token
938
tokens/s
prompt
354
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M3 4P+4E
RAM
16GB
OS
Darwin
Kernel Release
24.6.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.6.0: Mon Jan 19 22:01:13 PST 2026; root:xnu-11417.140.69.708.3~1/RELEASE_ARM64_T8122
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1116
tokens/s
74.9
tokens/s
930
ms
pp4096+tg256
817
tokens/s
47.6
tokens/s
5.03
sec
pp2048+tg256
997
tokens/s
62.1
tokens/s
2.07
sec
pp2048+tg768
1006
tokens/s
59.9
tokens/s
2.05
sec
pp1024+tg1024
1118
tokens/s
68.6
tokens/s
929
ms
pp1280+tg3072
1101
tokens/s
56.1
tokens/s
1.18
sec
pp384+tg1152
1121
tokens/s
74.9
tokens/s
355
ms
pp64+tg1024
853
tokens/s
80.3
tokens/s
86
ms
pp16+tg1536
309
tokens/s
76.9
tokens/s
63
ms