**问题描述** DeepSeek更新V4系列之后,模型的思考参数发生了变动,对齐了国际上主流使用的"reasoning_effort"参数; 目前软件中无法开启DeepSeek V4模型的 max 思考模式; <img width="586" height="427" alt="Image" src="https://github.com/user-attachments/assets/5f7ba695-6377-4bac-925e-c6ba1c075631" /> **附加上下文** 官方文档: https://api-docs.deepseek.com/zh-cn/guides/thinking_mode <img width="916" height="671" alt="Image" src="https://github.com/user-attachments/assets/e45c370b-4d53-4a5f-8e27-73d19385b4ae" />
问题描述
DeepSeek更新V4系列之后,模型的思考参数发生了变动,对齐了国际上主流使用的"reasoning_effort"参数;
目前软件中无法开启DeepSeek V4模型的 max 思考模式;
附加上下文
官方文档:
https://api-docs.deepseek.com/zh-cn/guides/thinking_mode