TEST #2759 RESULTS
02/19/2026 - 3:35 PM
ACCELERATOR
MODEL
11.1
tokens/s
generation
9.93
sec
time to first token
123
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
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 21:03:25 PST 2025; root:xnu-12377.61.12~1/RELEASE_ARM64_T8122
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
171
tokens/s
16.4
tokens/s
6.05
sec
pp4096+tg256
136
tokens/s
12.8
tokens/s
30.10
sec
pp2048+tg256
136
tokens/s
11.8
tokens/s
15.17
sec
pp2048+tg768
127
tokens/s
8.8
tokens/s
16.18
sec
pp1024+tg1024
134
tokens/s
10.3
tokens/s
7.69
sec
pp1280+tg3072
131
tokens/s
9.2
tokens/s
9.86
sec
pp384+tg1152
120
tokens/s
10.0
tokens/s
3.25
sec
pp64+tg1024
103
tokens/s
10.5
tokens/s
690
ms
pp16+tg1536
48
tokens/s
10.3
tokens/s
407
ms