Google unveils Genie 4, a AI model that creates realistic worlds with sound and music
deepmind google
| Source: Times Now on MSN | Original article
Google DeepMind introduced Genie 4, an AI model that generates photorealistic, interactive virtual worlds with real‑time audio from text and image prompts.
Google DeepMind unveiled Genie 4, its newest AI world model that can generate photorealistic, interactive virtual environments complete with real‑time physics and native audio. Users can start from text or image prompts, then explore, reshape and interact with the generated worlds as they evolve, hearing sound and music that respond to actions in the scene.
The launch marks a step up from earlier generative models that focused on static images or text. By coupling visual synthesis with dynamic physics simulation and on‑the‑fly audio, Genie 4 promises more immersive experiences for developers building games, simulations, or training tools. The model’s ability to render realistic sound alongside visual elements also opens possibilities for virtual production, education and remote collaboration, where auditory cues are essential for realism.
Google plans to make Genie 4 available first to developers and subscribers of its Google AI Ultra service, signalling a strategy of early access for power users while the broader public rollout remains pending. The move underscores Google’s broader push to embed generative AI deeper into creative workflows, complementing recent releases such as Gemini 4 Carbon for cybersecurity and AI‑enhanced search ads.
What to watch next: the range of third‑party applications that adopt Genie 4, pricing and usage limits for the AI Ultra tier, and how competitors respond in the emerging market for real‑time, audio‑aware world generation. Follow‑up details on performance benchmarks and integration tools are expected in the coming weeks.
Sources
Back to AIPULSEN