Google Unveils World Model Genie 4, Generating Visuals, Audio, and Dialogue in Real Time
Google DeepMind has released Genie 4, its world model capable of generating visuals, ambient sounds, and interactive dialogues in real time, while allowing users to modify scenes at any point using natural language. Powered by the Gemini family, the model employs Gemini 3.8 Flash as a guiding agent to break down multi-step tasks. It is now available as a research preview for trusted testers, with future access for Project Genie's Ultra subscribers and partners using the Gemini API and AI…
Google DeepMind has released Genie 4, a world model capable of generating visuals, ambient sounds, and interactive dialogues in real time, while also allowing users to modify scenes at any point using natural language.
The model is powered by the Gemini family, with Gemini 3.8 Flash serving as the guiding agent to break down multi-step tasks. It is now available as a research preview for trusted testers, and will later be extended to Ultra subscribers of Project Genie, as well as partners using the Gemini API and AI Studio. Generated content will be marked with a SynthID watermark.
Source: Google DeepMind (https://deepmind.google/blog/genie-4-a-new-generation-of-world-simulation/)
