Currently, if I set the editor to use GPT-4 in the chat window, it will set the CTRL-K inline editor to use GPT-4 as well.
I like to use the inline editor for simple tasks, and the Chat Window for more complex ones requiring a smarter AI. It would be a great QoL improvement, if the CTRL-K editor and Chat Window simply stored their last-used model in different places, because then I wouldn’t have to keep switching between models, and thus wouldn’t accidentally waste time and usage asking GPT-4 or Opus to re-indent the tabs of a 5-line function. That, or enable hotkeys to differentiate between querying the fastest model or the smartest model.
This has been annoying me for a while now, but now that I’ve added in the Claude models to my available models list, it makes going through the list more tedious.
