TEST #855 RESULTS
05/11/2025 - 4:19 PM
ACCELERATOR
MODEL
123
tokens/s
generation
567
ms
time to first token
2151
tokens/s
prompt
776
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Apple M4 Pro 10P+4E
RAM
24GB
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 20:42:08 PDT 2025; root:xnu-11417.121.6~4/RELEASE_ARM64_T6041
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2653
tokens/s
141
tokens/s
393
ms
pp4096+tg256
2193
tokens/s
79.9
tokens/s
1.88
sec
pp2048+tg256
2485
tokens/s
111
tokens/s
833
ms
pp2048+tg768
2477
tokens/s
106
tokens/s
836
ms
pp1024+tg1024
2566
tokens/s
126
tokens/s
406
ms
pp1280+tg3072
2571
tokens/s
98.9
tokens/s
505
ms
pp384+tg1152
2412
tokens/s
143
tokens/s
165
ms
pp64+tg1024
1544
tokens/s
156
tokens/s
48
ms
pp16+tg1536
454
tokens/s
148
tokens/s
41
ms