[image] 
官方给出的解释是： 
This is the first time we’ve seen two models perform comparably across the kinds of real-world coding tasks Augment handles every day. 
这是我们首次看到 两种模型在 Augment 每天处理的各种实际编码任务中表现相当 。 
We’re shipping a picker now because:
我们开启了picker功能，是因为： 

Sonnet 4 is fast and focused — great for quick edits and rapid iteration 
Sonnet 4 速度快、重点突出，非常适合快速编辑和快速迭代。
GPT-5 brings deeper reasoning and...