TEST #1401 RESULTS
08/20/2025 - 12:16 PM
ACCELERATOR
MODEL
78.8
tokens/s
generation
1.01
sec
time to first token
1204
tokens/s
prompt
455
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M1 Pro 8P+2E
RAM
16GB
OS
Darwin
Kernel Release
24.6.0
Architecture
arm64
Version
Cosmopolitan 3.9.7 MODE=aarch64; Darwin Kernel Version 24.6.0: Mon Jul 14 11:30:29 PDT 2025; root:xnu-11417.140.69~1/RELEASE_ARM64_T6000
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1458
tokens/s
88.7
tokens/s
714
ms
pp4096+tg256
1242
tokens/s
52.6
tokens/s
3.32
sec
pp2048+tg256
1399
tokens/s
71.4
tokens/s
1.48
sec
pp2048+tg768
1389
tokens/s
68.8
tokens/s
1.49
sec
pp1024+tg1024
1445
tokens/s
80.2
tokens/s
719
ms
pp1280+tg3072
1419
tokens/s
64.4
tokens/s
914
ms
pp384+tg1152
1422
tokens/s
90.4
tokens/s
280
ms
pp64+tg1024
799
tokens/s
98.7
tokens/s
89
ms
pp16+tg1536
257
tokens/s
94.0
tokens/s
71
ms