I sometimes want DeepSeek R1 to give me a short result without a long explanation, especially for routine questions. When I ask for “final answer only,” I’m not sure whether that changes how the model responds or simply leaves out the reasoning I would otherwise see.

For people who use R1 regularly, does requesting a concise final answer make the output more direct without hurting accuracy? I’m curious about your experience with everyday questions as well as tasks that need a bit of reasoning.