This article is an automated aggregation of coverage from news outlets. Oureditorialsoffer in-depth analysis curated by the NewsCord team.
Genie 4 adds native audio to real-time interactive world simulations. Users can steer scenes in real time using written prompts.
Save coverage image
Landscape · 1200 × 630Square · 1200 × 12004:3 · 1200 × 900Portrait · 1080 × 1350
Read the story
Genie 4’s interactive worlds
Google DeepMind announced Genie 4, a world model for real-time interactive simulation that it says brings “unprecedented realism and native audio” to simulations. Genie 4 lets users set an event in motion and steer scenes in real time using natural language, including “moving the camera, changing the weather, or introducing new characters.”
Genie 4 incorporates capabilities from the Gemini family, with “Gemini Omni anchors simulations in real-world appearance and behavior.” Genie 4 has started rolling out to a limited group of trusted testers, and it described coming access for “Google AI Ultra subscribers in Project Genie” and early access partners in the “Gemini API and AI Studio.”
Reporting for this sectionblog.google
Physics, coherence, and audio
Ars Technica reported that Genie 4 improves on Genie 3 in “almost every way,” and it described how a world model generates experiences frame by frame as users explore. Genie 4 uses Gemini 3.8 Flash as a guiding agent to break down goals and evaluate each output frame, helping keep simulations on track.
DeepMind notes that Genie 4 worlds “still only remain consistent for a couple of minutes before they forget how things started.” Genie 4 is capable of anticipating reaction, and it described the model as generating “music, ambient sounds and interactive dialogues” natively.
Reporting for this sectionArs TechnicaEuropa Press
Rollout, testers, and limits
TechCrunch said Genie 4 can maintain a consistent scene for up to two minutes, and it described the model as allowing users to “type prompts mid-simulation to change the scene.”Google warns that, as simulations run longer, “details and object permanence can shift over time.”
DeepMind characterizes Genie 4 as “its first real-time world model to generate audio natively,” and it described each world carrying ambient sound and interactive dialogue in real time. OODA Loop said Genie 4 can render explorable, physics-consistent digital worlds at “24 frames per second in high-definition 720p,” positioning the release as an expansion of real-time interactive 3D generation capabilities.
Reporting for this sectionTechCrunchThe Verge
