Feature Request: Autoparse of text in <think> tags in Thinking models

[xITmasterx]

This is a setting toggle that would be added to remove said think tags whenever one uses a thinking model like DeepSeek R1 or similar models, mainly to prevent their thinking output from spilling over to the actual chat.

[xITmasterx]

I would like to add that for reasons on the back-end, all models now have a switch to turn on or off thinking in selected models.

[stump-wiz]

To verify, are you saying that the LLM providers now allow you to disable the thinking output, so Writingway doesn't need to remove the thinking text?

[xITmasterx]

There would still be a need, given that older thinking models can't remove it. What I said was for the newer fusion thinking and normal models like Deepseek 3.1 or Kimi K2 0905.