TEST #1398 RESULTS
08/19/2025 - 11:50 PM
ACCELERATOR
MODEL
34.5
tokens/s
generation
7.70
sec
time to first token
179
tokens/s
prompt
93
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 5 6600H with Radeon Graphics (znver3)
RAM
38.7GB
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
210
tokens/s
38.6
tokens/s
4.91
sec
pp4096+tg256
161
tokens/s
27.8
tokens/s
25.50
sec
pp2048+tg256
175
tokens/s
32.8
tokens/s
11.70
sec
pp2048+tg768
184
tokens/s
31.9
tokens/s
11.19
sec
pp1024+tg1024
173
tokens/s
35.1
tokens/s
5.94
sec
pp1280+tg3072
171
tokens/s
30.3
tokens/s
7.50
sec
pp384+tg1152
184
tokens/s
37.0
tokens/s
2.11
sec
pp64+tg1024
183
tokens/s
38.8
tokens/s
374
ms
pp16+tg1536
171
tokens/s
38.1
tokens/s
119
ms