Google Gemini 3 is introduced as a new generation artificial intelligence model capable of creating websites, applications, and fully functional games based on a single prompt. The demonstrated capabilities of the new system indicate that this technology has advanced to a completely different level compared to its predecessors.
The editorial board reports that among the demonstrated examples of Gemini 3, a fully functional city-building simulator particularly stands out. The game is equipped with realistic road infrastructure, four types of residential and commercial spaces, money earning and spending mechanics, moving vehicles, and even a task system.
The second example created by the model is a physics-driven mini-Minecraft game. Here, the system generates voxel-structured objects with a single prompt, and then these objects are physically correctly destroyed. Users can test this directly in their browser without any additional applications.
Another demonstration is the ability to convert any image into an interactive voxel-based 3D scene. Furthermore, Gemini 3 can create an ideal business card website with a clean design, fully functional buttons, and optimized for all devices, all within seconds.
The immediate functionality of all these examples in the browser provides grounds to consider Gemini 3 as the beginning of new-generation artificial intelligence systems. It is stated that with its technological flexibility and generation quality, the model has created a new stage in the field of artificial intelligence.