TEST #1564 RESULTS
09/26/2025 - 2:53 PM
ACCELERATOR
MODEL
105
tokens/s
generation
325
ms
time to first token
12483
tokens/s
prompt
1595
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 8945HS w/ Radeon 780M Graphics (znver4)
RAM
30.6GB
OS
Linux
Kernel Release
6.16.3-76061603-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #202508231538~1758818838~24.04~0cd02b7 SMP PREEMPT_DYNAMIC Thu S
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
17423
tokens/s
113
tokens/s
67
ms
pp4096+tg256
12821
tokens/s
79.3
tokens/s
332
ms
pp2048+tg256
15344
tokens/s
98.3
tokens/s
144
ms
pp2048+tg768
15497
tokens/s
97.8
tokens/s
142
ms
pp1024+tg1024
17295
tokens/s
109
tokens/s
68
ms
pp1280+tg3072
15179
tokens/s
92.6
tokens/s
93
ms
pp384+tg1152
15504
tokens/s
117
tokens/s
33
ms
pp64+tg1024
3280
tokens/s
123
tokens/s
28
ms
pp16+tg1536
8
tokens/s
120
tokens/s
2.01
sec