TEST #1929 RESULTS
12/11/2025 - 6:04 PM
ACCELERATOR
MODEL
14.7
tokens/s
generation
8.67
sec
time to first token
136
tokens/s
prompt
61
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 4P+6E
RAM
24GB
OS
Darwin
Kernel Release
25.2.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.2.0: Thu Nov 13 23:28:51 PST 2025; root:xnu-12377.61.11~7/RELEASE_ARM64_T8132
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
178
tokens/s
18.1
tokens/s
5.82
sec
pp4096+tg256
155
tokens/s
13.4
tokens/s
26.45
sec
pp2048+tg256
158
tokens/s
16.4
tokens/s
12.99
sec
pp2048+tg768
161
tokens/s
15.1
tokens/s
12.81
sec
pp1024+tg1024
161
tokens/s
15.1
tokens/s
6.42
sec
pp1280+tg3072
147
tokens/s
12.9
tokens/s
8.74
sec
pp384+tg1152
104
tokens/s
13.4
tokens/s
3.75
sec
pp64+tg1024
105
tokens/s
14.1
tokens/s
695
ms
pp16+tg1536
58
tokens/s
13.6
tokens/s
335
ms