TEST #1583 RESULTS
10/05/2025 - 8:38 AM
ACCELERATOR
MODEL
78.6
tokens/s
generation
1.16
sec
time to first token
1096
tokens/s
prompt
420
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 4P+6E
RAM
16GB
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 19:54:43 PDT 2025; root:xnu-11417.121.6~2/RELEASE_ARM64_T8132
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1345
tokens/s
87.3
tokens/s
773
ms
pp4096+tg256
1006
tokens/s
56.3
tokens/s
4.09
sec
pp2048+tg256
1224
tokens/s
72.6
tokens/s
1.69
sec
pp2048+tg768
1220
tokens/s
70.0
tokens/s
1.69
sec
pp1024+tg1024
1313
tokens/s
80.3
tokens/s
791
ms
pp1280+tg3072
1293
tokens/s
66.5
tokens/s
1.00
sec
pp384+tg1152
1369
tokens/s
88.4
tokens/s
291
ms
pp64+tg1024
822
tokens/s
94.7
tokens/s
88
ms
pp16+tg1536
271
tokens/s
90.9
tokens/s
69
ms