Google DeepMind released Genie 4, its newest world model for real-time interactive simulation, as a research preview on Tuesday, according to a DeepMind post, TechCrunch and The Verge.
Google said Genie 4 is natively multimodal, generating music, ambient noises and interactive dialogue with the simulated world. Users can steer scenes in real time with natural-language prompts, such as changing weather, moving the camera or adding people and objects, DeepMind said. The Verge reported Google says experiences can run about two minutes, up from about one minute on Genie 3. Specs and capability claims are attributed to Google and have not been independently verified.
DeepMind said Genie 4 is rolling out first as a research preview for trusted testers. Google said it is coming soon to Google AI Ultra subscribers in Project Genie and to early-access partners of the Gemini API and Google AI Studio. To maintain transparency, DeepMind said all video and audio generated by Genie 4 will include SynthID, its watermarking technology for AI-generated content.
Bloomberg described Genie as a technology that has unsettled video game investors. When Google unveiled Project Genie, a prototype built on Genie 3, in January, video game stocks fell sharply as investors weighed whether AI tools could change how games are made, Bloomberg reported. Asked at a press briefing about the impact on the games industry, Google DeepMind product manager Diego Rivas said the company is working with partners to understand how the technology fits into their work and where its limits are, according to Bloomberg.
Google also listed known gaps. DeepMind said visual details and object permanence can drift as a simulation runs longer, and that complex physical interactions, real-time audio accuracy, rendering legible text, spatial accuracy and long-term consistency remain active research problems. TechCrunch similarly reported the team has not fully solved audio accuracy or legible in-world text. Those caveats sit alongside Google’s claim that Genie 4 is more realistic and more steerable than prior Genie releases.
Sources
- Google DeepMind: Genie 4 — A new generation of world simulation (October 27, 2026)
- TechCrunch (Rebecca Bellan): Google’s Genie 4 world model adds audio, more realistic worlds, and real-time steering (October 27, 2026)
- The Verge (Jay Peters): Google’s next Genie AI model generates more realistic experiences (October 27, 2026)
- Bloomberg (Davey Alba): Google upgrades AI world model that rattled video game stocks (October 27, 2026)
- Bloomberg Línea (Spanish edition of the Bloomberg story) (October 27, 2026)