TEST #862 RESULTS
05/14/2025 - 8:04 AM
ACCELERATOR
MODEL
88.2
tokens/s
generation
952
ms
time to first token
1323
tokens/s
prompt
497
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M3 Pro 5P+6E
RAM
36GB
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 19:54:29 PDT 2025; root:xnu-11417.121.6~2/RELEASE_ARM64_T6030
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1645
tokens/s
98.0
tokens/s
633
ms
pp4096+tg256
1235
tokens/s
62.9
tokens/s
3.33
sec
pp2048+tg256
1502
tokens/s
81.7
tokens/s
1.38
sec
pp2048+tg768
1491
tokens/s
78.7
tokens/s
1.39
sec
pp1024+tg1024
1633
tokens/s
89.8
tokens/s
637
ms
pp1280+tg3072
1613
tokens/s
74.7
tokens/s
805
ms
pp384+tg1152
1523
tokens/s
99.1
tokens/s
261
ms
pp64+tg1024
992
tokens/s
106
tokens/s
73
ms
pp16+tg1536
272
tokens/s
103
tokens/s
67
ms