We tested 7 AI models (including Gemini 3 Pro Preview released yesterday and Grok 4.1, which was released a day before Gemini 3) for a semi-complicated UI task: Build a beautiful and functional dashboard using sample data.
Somebody is clearly blind Codex’ ui is objectively the best looking out of all of them! 🤯
But I guess the new shiny object (Gemini 3) syndrome is kicking in hard😎
Somebody is clearly blind Codex’ ui is objectively the best looking out of all of them! 🤯
But I guess the new shiny object (Gemini 3) syndrome is kicking in hard😎