TEST #2320 RESULTS
01/17/2026 - 6:11 AM
ACCELERATOR
MODEL
70.0
tokens/s
generation
3.08
sec
time to first token
452
tokens/s
prompt
217
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 9 275HX (arrowlake-s)
RAM
63.4GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
485
tokens/s
72.4
tokens/s
2.13
sec
pp4096+tg256
391
tokens/s
63.3
tokens/s
10.49
sec
pp2048+tg256
447
tokens/s
70.4
tokens/s
4.60
sec
pp2048+tg768
445
tokens/s
68.8
tokens/s
4.62
sec
pp1024+tg1024
473
tokens/s
69.3
tokens/s
2.18
sec
pp1280+tg3072
474
tokens/s
64.8
tokens/s
2.71
sec
pp384+tg1152
491
tokens/s
72.9
tokens/s
796
ms
pp64+tg1024
493
tokens/s
74.6
tokens/s
143
ms
pp16+tg1536
371
tokens/s
73.9
tokens/s
57
ms