GPT 5.6-Sol 模型配置model.json中新增UltraFast等级，描述为： 

The fastest available responses for latency-sensitive work. 

 [image]