TEST #617 RESULTS
04/08/2025 - 9:07 PM
ACCELERATOR
MODEL
176
tokens/s
generation
334
ms
time to first token
3296
tokens/s
prompt
1203
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M2 Ultra 16P+8E
RAM
128GB
OS
Darwin
Kernel Release
24.3.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.3.0: Thu Jan 2 20:24:23 PST 2025; root:xnu-11215.81.4~3/RELEASE_ARM64_T6020
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4077
tokens/s
194
tokens/s
257
ms
pp4096+tg256
4324
tokens/s
127
tokens/s
955
ms
pp2048+tg256
4474
tokens/s
165
tokens/s
464
ms
pp2048+tg768
4457
tokens/s
159
tokens/s
466
ms
pp1024+tg1024
4165
tokens/s
181
tokens/s
251
ms
pp1280+tg3072
4166
tokens/s
150
tokens/s
313
ms
pp384+tg1152
2993
tokens/s
197
tokens/s
133
ms
pp64+tg1024
787
tokens/s
211
tokens/s
86
ms
pp16+tg1536
216
tokens/s
204
tokens/s
78
ms