TEST #3548 RESULTS
05/16/2026 - 8:37 AM
ACCELERATOR
MODEL
186
tokens/s
generation
119
ms
time to first token
11428
tokens/s
prompt
2612
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i9-14900HX (alderlake)
RAM
15.7GB
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
15178
tokens/s
184
tokens/s
74
ms
pp4096+tg256
9780
tokens/s
180
tokens/s
425
ms
pp2048+tg256
12772
tokens/s
179
tokens/s
166
ms
pp2048+tg768
12753
tokens/s
188
tokens/s
167
ms
pp1024+tg1024
13735
tokens/s
201
tokens/s
81
ms
pp1280+tg3072
14098
tokens/s
183
tokens/s
96
ms
pp384+tg1152
16244
tokens/s
189
tokens/s
31
ms
pp64+tg1024
6782
tokens/s
192
tokens/s
17
ms
pp16+tg1536
1507
tokens/s
180
tokens/s
18
ms