TEST #3286 RESULTS
04/11/2026 - 9:21 AM
ACCELERATOR
MODEL
19.6
tokens/s
generation
6.61
sec
time to first token
177
tokens/s
prompt
81
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M5 4P+6E
RAM
32GB
OS
Darwin
Kernel Release
25.4.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 25.4.0: Thu Mar 19 19:33:43 PDT 2026; root:xnu-12377.101.15~1/RELEASE_ARM64_T8142
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
228
tokens/s
23.8
tokens/s
4.53
sec
pp4096+tg256
207
tokens/s
19.3
tokens/s
19.87
sec
pp2048+tg256
213
tokens/s
21.8
tokens/s
9.68
sec
pp2048+tg768
209
tokens/s
21.0
tokens/s
9.86
sec
pp1024+tg1024
203
tokens/s
21.7
tokens/s
5.10
sec
pp1280+tg3072
177
tokens/s
16.6
tokens/s
7.27
sec
pp384+tg1152
162
tokens/s
17.8
tokens/s
2.42
sec
pp64+tg1024
142
tokens/s
17.7
tokens/s
494
ms
pp16+tg1536
57
tokens/s
16.7
tokens/s
323
ms