Is there a way to hide what's between <think></think> for reasoning LLM models?

[miriameng]

For reasoning models, now it by default shows the thinking process of the LLM with the tag, is there a way to hide it with a setting, or perhaps even collapse it if ST supports it?

[yaroslavyaroslav]

Nope it's not implemented yet. Although there is a foundation for that in rust's part of this plugin. I mean that all the thinking part is stored separately in cache from the rest answer. But it still has to be pulled through the whole chain and I don't I'd have time to do that in this quarter unfortunately.

UPD: there's literally no way to split it up in the broadcasting part in an easy reliable way without writing your own parser on the python's side, so it could be quite hacky to do that well