TEST #2891 RESULTS
02/24/2026 - 8:27 AM
ACCELERATOR
MODEL
15.3
tokens/s
generation
9.18
sec
time to first token
134
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
16GB
OS
Darwin
Kernel Release
25.2.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.2.0: Tue Nov 18 21:08:48 PST 2025; root:xnu-12377.61.12~1/RELEASE_ARM64_T8132
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
159
tokens/s
17.3
tokens/s
6.50
sec
pp4096+tg256
144
tokens/s
13.6
tokens/s
28.45
sec
pp2048+tg256
152
tokens/s
15.8
tokens/s
13.54
sec
pp2048+tg768
146
tokens/s
15.4
tokens/s
14.05
sec
pp1024+tg1024
145
tokens/s
16.1
tokens/s
7.13
sec
pp1280+tg3072
142
tokens/s
14.0
tokens/s
9.10
sec
pp384+tg1152
131
tokens/s
15.3
tokens/s
2.99
sec
pp64+tg1024
129
tokens/s
15.6
tokens/s
559
ms
pp16+tg1536
58
tokens/s
14.5
tokens/s
335
ms