TEST #2579 RESULTS
02/10/2026 - 9:11 AM
ACCELERATOR
MODEL
68.4
tokens/s
generation
505
ms
time to first token
2324
tokens/s
prompt
680
LocalScore
HOW YOU STACK UP
Explore All ResultsNeo 3 3B A400M Base - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-14450HX (alderlake)
RAM
15.7GB
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
3167
tokens/s
60.2
tokens/s
337
ms
pp4096+tg256
2592
tokens/s
71.9
tokens/s
1.59
sec
pp2048+tg256
2873
tokens/s
67.4
tokens/s
726
ms
pp2048+tg768
3012
tokens/s
73.2
tokens/s
692
ms
pp1024+tg1024
3119
tokens/s
67.3
tokens/s
346
ms
pp1280+tg3072
2776
tokens/s
66.8
tokens/s
475
ms
pp384+tg1152
2533
tokens/s
69.7
tokens/s
167
ms
pp64+tg1024
648
tokens/s
67.7
tokens/s
114
ms
pp16+tg1536
196
tokens/s
71.5
tokens/s
98
ms