TEST #2307 RESULTS
01/15/2026 - 6:41 PM
ACCELERATOR
MODEL
90.1
tokens/s
generation
219
ms
time to first token
5964
tokens/s
prompt
1349
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen3 1.7B - Q8_0
SYSTEM
CPU
AMD Ryzen 5 5600 6-Core Processor (znver3)
RAM
31.9GB
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
7605
tokens/s
88.9
tokens/s
146
ms
pp4096+tg256
5608
tokens/s
87.5
tokens/s
743
ms
pp2048+tg256
6808
tokens/s
92.3
tokens/s
312
ms
pp2048+tg768
6802
tokens/s
92.6
tokens/s
314
ms
pp1024+tg1024
7575
tokens/s
90.1
tokens/s
146
ms
pp1280+tg3072
7327
tokens/s
90.4
tokens/s
186
ms
pp384+tg1152
7949
tokens/s
90.8
tokens/s
60
ms
pp64+tg1024
3192
tokens/s
88.7
tokens/s
33
ms
pp16+tg1536
808
tokens/s
89.2
tokens/s
31
ms