TEST #1225 RESULTS
07/16/2025 - 1:28 AM
ACCELERATOR
MODEL
81.6
tokens/s
generation
416
ms
time to first token
3624
tokens/s
prompt
893
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 7 155H (alderlake)
RAM
95.5GB
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
4569
tokens/s
86.8
tokens/s
235
ms
pp4096+tg256
2558
tokens/s
79.5
tokens/s
1.61
sec
pp2048+tg256
3526
tokens/s
83.8
tokens/s
593
ms
pp2048+tg768
3576
tokens/s
81.7
tokens/s
585
ms
pp1024+tg1024
4349
tokens/s
83.2
tokens/s
247
ms
pp1280+tg3072
4072
tokens/s
77.8
tokens/s
325
ms
pp384+tg1152
5643
tokens/s
80.2
tokens/s
79
ms
pp64+tg1024
3130
tokens/s
79.4
tokens/s
35
ms
pp16+tg1536
1196
tokens/s
81.9
tokens/s
25
ms