[Discussion]: Deepseek-v4-flash api close thinking model · CherryHQ/cherry-studio#15199

2026-05-20T02:26:59.000Z

### Issue Checklist - [x] I understand that issues are meant for feedback and problem-solving, not for venting, and I will provide as much detail as possible to help resolve the issue. - [x] I have checked the pinned issues and searched through the existing [open issues](https://github.com/CherryHQ/cherry-studio/issues), [closed issues](https://github.com/CherryHQ/cherry-studio/issues?q=is%3Aissue%20state%3Aclosed), and [discussions](https://github.com/CherryHQ/cherry-studio/discussions) and did not find a similar suggestion. - [x] I confirm that I am here to ask questions and discuss issues, not to report bugs or request features. ### Platform Windows ### Version 1.9.6 ### Your Question When I call the deepseek-v4-flash API in Cherry Studio, I use some agents for translation and paper reading. However, the model enters thinking mode every time. For these use cases, I do not need thinking mode. How can I disable it? I have tried turning off “Thinking” in the Advanced Settings of the Cherry Studio interface, but it does not seem to take effect. So I would like to report this issue and hope it can be fixed in the next version update. Thank you. ### Context _No response_ ### Additional Information ```shell ``` ### Priority Medium (Would like a response soon)

(0 留言) (1 反應) (0 負責人)TypeScript (45,688 star) (4,338 fork)batch import

Modelsdiscussionhelp wanted

描述

Issue Checklist

I understand that issues are meant for feedback and problem-solving, not for venting, and I will provide as much detail as possible to help resolve the issue.
I have checked the pinned issues and searched through the existing open issues, closed issues, and discussions and did not find a similar suggestion.
I confirm that I am here to ask questions and discuss issues, not to report bugs or request features.

Platform

Windows

Version

1.9.6

Your Question

When I call the deepseek-v4-flash API in Cherry Studio, I use some agents for translation and paper reading. However, the model enters thinking mode every time. For these use cases, I do not need thinking mode. How can I disable it? I have tried turning off “Thinking” in the Advanced Settings of the Cherry Studio interface, but it does not seem to take effect. So I would like to report this issue and hope it can be fixed in the next version update. Thank you.

Context

No response

Additional Information

Priority

Medium (Would like a response soon)

貢獻者指南

技術棧: typescriptreact
領域: frontend
議題類型: bug
難度: 2
預計時間: 1-3 hours
活動狀態: fresh
清晰度: clear
前置要求: Basic understanding of ReactFamiliarity with API integration in TypeScript
新手友善度: 65
研究方向: Investigate the 'Advanced Settings' component in the Cherry Studio frontend. Look for where the 'Thinking' toggle state is managed and how it is passed to API call parameters for DeepSeek models. Check the API integration file for DeepSeek (likely in src/providers or similar) to see if the toggle's value is being read or if a parameter to disable thinking is missing. Review any recent changes that might have broken this functionality.