TEST #4134 RESULTS
09/09/2026 - 8:00 AM
ACCELERATOR
MODEL
63.8
tokens/s
generation
2.42
sec
time to first token
557
tokens/s
prompt
245
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen AI 9 HX 370 w/ Radeon 890M
RAM
61.4GB
OS
Linux
Kernel Release
7.0.0-111030-tuxedo
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #30~24.04.1tux1 SMP PREEMPT_DYNAMIC Tue Sep 1 19:13:45 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
614
tokens/s
71.0
tokens/s
1.68
sec
pp4096+tg256
492
tokens/s
53.1
tokens/s
8.34
sec
pp2048+tg256
590
tokens/s
62.5
tokens/s
3.49
sec
pp2048+tg768
553
tokens/s
60.2
tokens/s
3.72
sec
pp1024+tg1024
627
tokens/s
64.7
tokens/s
1.65
sec
pp1280+tg3072
631
tokens/s
58.7
tokens/s
2.04
sec
pp384+tg1152
608
tokens/s
65.7
tokens/s
645
ms
pp64+tg1024
519
tokens/s
69.9
tokens/s
138
ms
pp16+tg1536
382
tokens/s
68.7
tokens/s
55
ms