TEST #1293 RESULTS
08/03/2025 - 8:51 AM
ACCELERATOR
MODEL
44.7
tokens/s
generation
2.49
sec
time to first token
474
tokens/s
prompt
204
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M2 Max 8P+4E
RAM
32GB
OS
Darwin
Kernel Release
24.5.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.5.0: Tue Apr 22 19:54:25 PDT 2025; root:xnu-11417.121.6~2/RELEASE_ARM64_T6020
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
568
tokens/s
49.3
tokens/s
1.82
sec
pp4096+tg256
537
tokens/s
35.0
tokens/s
7.66
sec
pp2048+tg256
561
tokens/s
42.6
tokens/s
3.68
sec
pp2048+tg768
561
tokens/s
41.6
tokens/s
3.68
sec
pp1024+tg1024
509
tokens/s
46.0
tokens/s
2.03
sec
pp1280+tg3072
531
tokens/s
40.0
tokens/s
2.43
sec
pp384+tg1152
544
tokens/s
49.7
tokens/s
725
ms
pp64+tg1024
339
tokens/s
50.8
tokens/s
207
ms
pp16+tg1536
114
tokens/s
46.8
tokens/s
158
ms