Some people say that the thinking slows the system down for no real reason.
Thinking to me seems like a “to do” list kind of what Claude Code or Codex does. Maybe thinking is better with the AI in a harness that creates this to do list and doesn’t rely solely on the model.
And if I want to play with this, i can’t find a way to shut of thinking on LM Studio for this model on my Mac.
[link] [comments]