[ENHANCEMENT] we should have a choice on “thinking enable”when choose Z AI as model provider

[BeWater799]

Problem (one or two sentences)

as i mentioned we dont have a choice when using officially provided GLM4.6

Context (who is affected and when)

as i mentioned above.

Desired behavior (conceptual, not technical)

add a bottom to choose enable thinking or not when make an API setting

Constraints / preferences (optional)

No response

Request checklist

• I've searched existing Issues and Discussions for duplicates
• This describes a specific problem with clear context and impact

Roo Code Task Links (optional)

No response

Acceptance criteria (optional)

No response

Proposed approach (optional)

if users can switch thinking mode when tasking is better,this could be important, models like deepseek v3.2 and GLM4.6 could have different features ，capabilities and application scenarios when the thinking mode is on and of.we could even develop an func to enable models to ask to switch itself like use"sequantial thinking" .

Trade-offs / risks (optional)

No response

[roomote]

I'm exploring a potential fix for adding a thinking mode toggle for Z AI provider models like GLM4.6 and DeepSeek v3.2. I'll open a PR to share progress.

[roomote]

I've opened PR #8466 with my implementation of the thinking mode toggle for Z AI provider models.

The implementation adds a checkbox in the Z AI settings that allows users to enable/disable thinking mode for models like GLM-4.6 and DeepSeek v3.2. The feature is opt-in and defaults to disabled for backward compatibility.

Feedback and suggestions are welcome!