TEST #372 RESULTS
04/06/2025 - 2:00 PM
ACCELERATOR
MODEL
31.8
tokens/s
generation
3.79
sec
time to first token
329
tokens/s
prompt
140
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 Pro 10P+4E
RAM
64GB
OS
Darwin
Kernel Release
24.4.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.4.0: Wed Mar 12 21:24:54 PDT 2025; root:xnu-11417.101.13~4/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
398
tokens/s
35.0
tokens/s
2.60
sec
pp4096+tg256
365
tokens/s
24.6
tokens/s
11.28
sec
pp2048+tg256
357
tokens/s
30.3
tokens/s
5.77
sec
pp2048+tg768
329
tokens/s
28.9
tokens/s
6.26
sec
pp1024+tg1024
340
tokens/s
31.9
tokens/s
3.04
sec
pp1280+tg3072
343
tokens/s
28.0
tokens/s
3.77
sec
pp384+tg1152
372
tokens/s
35.1
tokens/s
1.06
sec
pp64+tg1024
331
tokens/s
36.7
tokens/s
219
ms
pp16+tg1536
129
tokens/s
35.8
tokens/s
149
ms