TEST #1173 RESULTS
07/02/2025 - 3:49 PM
ACCELERATOR
MODEL
89.9
tokens/s
generation
803
ms
time to first token
1616
tokens/s
prompt
566
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M3 Pro 6P+6E
RAM
36GB
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 19:54:29 PDT 2025; root:xnu-11417.121.6~2/RELEASE_ARM64_T6030
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2016
tokens/s
99.1
tokens/s
518
ms
pp4096+tg256
1427
tokens/s
63.7
tokens/s
2.89
sec
pp2048+tg256
1803
tokens/s
83.3
tokens/s
1.15
sec
pp2048+tg768
1791
tokens/s
80.1
tokens/s
1.16
sec
pp1024+tg1024
1968
tokens/s
92.1
tokens/s
530
ms
pp1280+tg3072
1932
tokens/s
76.1
tokens/s
673
ms
pp384+tg1152
1946
tokens/s
101
tokens/s
206
ms
pp64+tg1024
1331
tokens/s
108
tokens/s
56
ms
pp16+tg1536
331
tokens/s
105
tokens/s
57
ms