CODE415 RADAR ·
DEMOS
· SEP 26, 2026
EN
中文
日本語
@Code415zg
Can your laptop run GLM-5.2?
Warming up…
Model
GLM-5.2
Kimi K3
DeepSeek V4 Flash
Qwen3.6
View
Robots
Technical
Drag to rotate
RAM
16 GB
24 GB
32 GB
64 GB
128 GB
256 GB
SSD
0.5 GB/s
3 GB/s
6 GB/s
12 GB/s
Prefetch
On
Off
Speed
1×
10×
Cold start
Speed, estimated
…
Found in RAM
…
Read from SSD
…
Experts kept in RAM
…
Time per token
…
What the model writes, at this speed
What you are looking at
The four models
Three tricks that make it bearable
Running it yourself or calling an API
How this page works