Which version of the app are you using?
v3.3.21
Which API Provider are you using?
Ollama
Which Model are you using?
deepseek-r1:*
What happened?
The content of the think tag is mixed with the API response, making it hard to distinguish.
It could also lead to a significant increase in context. I'm considering whether some OPENAI-compatible deekseek providers might output tags in their responses, and if so, this patch might need to be applied to other providers as well.
I’ve tested that LMStudio removes the think tag before responding.

Steps to reproduce

Relevant API REQUEST output
Additional context
The content of think tag should be shown in reasoning block.
Patch: #1080
