TEST #1052 RESULTS
06/14/2025 - 2:57 PM
ACCELERATOR
MODEL
26.3
tokens/s
generation
4.71
sec
time to first token
262
tokens/s
prompt
113
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M2 Pro 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
296
tokens/s
28.5
tokens/s
3.49
sec
pp4096+tg256
274
tokens/s
20.8
tokens/s
15.00
sec
pp2048+tg256
291
tokens/s
25.1
tokens/s
7.08
sec
pp2048+tg768
291
tokens/s
24.4
tokens/s
7.08
sec
pp1024+tg1024
298
tokens/s
26.8
tokens/s
3.48
sec
pp1280+tg3072
295
tokens/s
23.6
tokens/s
4.38
sec
pp384+tg1152
295
tokens/s
28.6
tokens/s
1.33
sec
pp64+tg1024
227
tokens/s
29.9
tokens/s
313
ms
pp16+tg1536
89
tokens/s
28.8
tokens/s
211
ms