TEST #1466 RESULTS
08/31/2025 - 12:01 PM
ACCELERATOR
MODEL
0.3
tokens/s
generation
2959.60
sec
time to first token
1
tokens/s
prompt
0
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Mobile AMD Sempron(tm) Processor 3400+
RAM
2.7GB
OS
Linux
Kernel Release
5.10.10-64
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #3 SMP PREEMPT Sun Feb 7 09:26:54 CET 2021
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
1
tokens/s
0.3
tokens/s
1999.00
sec
pp4096+tg256
0
tokens/s
0.2
tokens/s
10455.65
sec
pp2048+tg256
0
tokens/s
0.3
tokens/s
4404.84
sec
pp2048+tg768
0
tokens/s
0.3
tokens/s
4402.67
sec
pp1024+tg1024
1
tokens/s
0.3
tokens/s
1997.62
sec
pp1280+tg3072
1
tokens/s
0.3
tokens/s
2544.48
sec
pp384+tg1152
1
tokens/s
0.4
tokens/s
692.52
sec
pp64+tg1024
1
tokens/s
0.4
tokens/s
109.94
sec
pp16+tg1536
1
tokens/s
0.4
tokens/s
29.67
sec