TEST #3647 RESULTS
05/29/2026 - 10:50 AM
ACCELERATOR
MODEL
21.4
tokens/s
generation
12.85
sec
time to first token
115
tokens/s
prompt
58
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-8600K CPU @ 3.60GHz (skylake)
RAM
4GB
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
120
tokens/s
23.4
tokens/s
8.55
sec
pp4096+tg256
89
tokens/s
17.4
tokens/s
45.88
sec
pp2048+tg256
109
tokens/s
20.7
tokens/s
18.87
sec
pp2048+tg768
108
tokens/s
20.0
tokens/s
18.99
sec
pp1024+tg1024
120
tokens/s
22.1
tokens/s
8.61
sec
pp1280+tg3072
118
tokens/s
19.3
tokens/s
10.93
sec
pp384+tg1152
127
tokens/s
23.2
tokens/s
3.07
sec
pp64+tg1024
130
tokens/s
23.5
tokens/s
527
ms
pp16+tg1536
117
tokens/s
23.1
tokens/s
179
ms