TEST #590 RESULTS
04/08/2025 - 11:47 AM
ACCELERATOR
MODEL
10.8
tokens/s
generation
9.36
sec
time to first token
121
tokens/s
prompt
52
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M3 4P+4E
RAM
24GB
OS
Darwin
Kernel Release
24.4.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.4.0: Wed Mar 19 21:18:00 PDT 2025; root:xnu-11417.101.15~1/RELEASE_ARM64_T8122
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
176
tokens/s
16.7
tokens/s
5.86
sec
pp4096+tg256
152
tokens/s
13.4
tokens/s
27.09
sec
pp2048+tg256
149
tokens/s
15.0
tokens/s
13.86
sec
pp2048+tg768
142
tokens/s
11.8
tokens/s
14.54
sec
pp1024+tg1024
136
tokens/s
11.6
tokens/s
7.61
sec
pp1280+tg3072
143
tokens/s
8.4
tokens/s
9.04
sec
pp384+tg1152
83
tokens/s
6.7
tokens/s
4.70
sec
pp64+tg1024
74
tokens/s
6.7
tokens/s
962
ms
pp16+tg1536
37
tokens/s
6.5
tokens/s
536
ms