TEST #1080 RESULTS
06/20/2025 - 12:16 AM
ACCELERATOR
MODEL
205
tokens/s
generation
113
ms
time to first token
11930
tokens/s
prompt
2787
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5800X3D 8-Core Processor (znver3)
RAM
63.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
15820
tokens/s
213
tokens/s
69
ms
pp4096+tg256
10390
tokens/s
200
tokens/s
399
ms
pp2048+tg256
13453
tokens/s
212
tokens/s
157
ms
pp2048+tg768
13439
tokens/s
217
tokens/s
157
ms
pp1024+tg1024
15802
tokens/s
203
tokens/s
69
ms
pp1280+tg3072
14698
tokens/s
200
tokens/s
97
ms
pp384+tg1152
16456
tokens/s
202
tokens/s
28
ms
pp64+tg1024
6466
tokens/s
192
tokens/s
19
ms
pp16+tg1536
851
tokens/s
212
tokens/s
25
ms