TEST #3363 RESULTS
04/22/2026 - 4:11 PM
ACCELERATOR
MODEL
9.7
tokens/s
generation
12.88
sec
time to first token
101
tokens/s
prompt
42
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M1 4P+4E
RAM
16GB
OS
Darwin
Kernel Release
25.3.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.3.0: Wed Jan 28 20:53:01 PST 2026; root:xnu-12377.81.4~5/RELEASE_ARM64_T8103
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
112
tokens/s
10.7
tokens/s
9.24
sec
pp4096+tg256
98
tokens/s
8.1
tokens/s
42.10
sec
pp2048+tg256
107
tokens/s
9.3
tokens/s
19.29
sec
pp2048+tg768
106
tokens/s
9.0
tokens/s
19.47
sec
pp1024+tg1024
110
tokens/s
9.8
tokens/s
9.39
sec
pp1280+tg3072
109
tokens/s
9.0
tokens/s
11.84
sec
pp384+tg1152
113
tokens/s
10.4
tokens/s
3.47
sec
pp64+tg1024
109
tokens/s
10.7
tokens/s
677
ms
pp16+tg1536
48
tokens/s
10.5
tokens/s
422
ms