I compared GPT-5.1 to GPT-5 on ChatGPT, and now I don’t want to go back

Some ChatGPT users greeted the introduction of GPT-5.1 as the default model for the chatbot with some skepticism – after all, more than a few users vocally complained only a few months ago when GPT-5 debuted. But there's been no comparable furor this time around, probably because GPT-5.1 seems calculated to address those earlier complaints, which might never have come up if GPT-5 had anticipated them.

GPT-5.1 isn't claimed to be a quantum leap in AI as GPT-5 was, but it does upgrade some key elements of the AI model, and ultimately is more responsive and better at imitating a pleasant personality. That much was clear from some impromptu use. I then decided to do a slightly more formal test of the areas where OpenAI claimed GPT-5.1 outperformed GPT-5.

I ran prompts crafted to test those upgraded capabilities through both GPT-5.1 ...