Google updates Genie AI model to support two-minute interactive simulations
Google has launched Genie 4, an upgraded AI model capable of generating two-minute interactive simulations with integrated audio and real-time text-based editing.
Google released its latest iteration of the Genie AI model, designed to build interactive environments from text or image prompts. This version, Genie 4, extends the maximum duration of generated simulations to two minutes, doubling the capacity of its predecessor.
Technically, the update integrates Google’s Gemini 3.8 Flash model to manage the underlying simulation logic. Key functional additions include the ability to generate synchronized audio and the capacity for users to modify environments in real-time through natural language commands.
The company suggests potential applications for the tool in creative and industrial sectors, specifically naming film pre-visualization, gaming prototypes, and robotic training scenarios as target use cases. By allowing users to iterate on 3D environments through textual instructions, the model aims to reduce the time required for early-stage design and prototyping.