Google DeepMind began rolling out Genie 4, its newest "world model" for generating interactive digital environments, to eligible Google AI Ultra subscribers on 18 November 2026. Diego Rivas, Group Product Manager at Google DeepMind, and Pavan Kumar, Group Product Manager at Google Labs, announced the launch in a post on Google's official blog. It comes as part of a revamped Project Genie experience.
What is Genie 4, and what can users do with it?
Genie 4 lets users create a virtual world from a prompt, walk through it in a web browser, and change it while they explore. Google calls it its "most realistic and interactive world model yet". Users can type a new instruction mid-session to change the weather, add a character or redirect the action. For the first time, Google says, the visuals come with real-time native audio that changes with the scene, from background sounds to character dialogue.
Project Genie began earlier this year as an experimental research prototype. Google says early testers asked for more control over what happens inside these worlds and easier ways to build and share them.
What changes for Genie 3?
The older Genie 3 model stays available in Project Genie's model picker. Google said Genie 3 is the option for longer runs or worlds built from panoramas, and listed these updates:
- sessions of up to three minutes per generation, up from 60 seconds, plus the option to chain several worlds into one longer journey;
- new worlds seeded from 360-degree panorama images;
- real-time, prompt-driven events that change the environment during exploration.
A creation agent and a community gallery
A new Project Genie agent lets users describe a world in plain language. It then calls on Google's multimodal models to design the world's look and compose matching background music. Gemini models can also join a session: they watch what is on screen, suggest where to go next, or trigger in-world events.
A community gallery mixes curated worlds with user creations that others can explore and remix. Google highlighted collaborations with early-access creators, including "Villa Giallo", a room-by-room mystery in a gothic manor made with creative technologist Will aka BLVCKL!GHT, and "The Life of Ants", built with filmmaker Anand Gandhi, which lets players move through an ant colony from an ant's point of view.
What safeguards has Google described?
Google says all video and audio generated in Project Genie carries SynthID watermarking, and all model inputs and outputs go through safety screening. The company describes Genie 3 and Genie 4 as frontier research. It names camera latency, complex physical interactions, precise real-time audio and keeping scenes consistent over long periods as problems it is still working on.
Who gets it next?
Only eligible Google AI Ultra subscribers can use the new experience at launch. Google AI Pro members can join a waitlist. The post does not say which countries are included. Earlier this week, MapMyTimes reported that Google made Gemini's Tasks mode free in six countries, including India.
Photo source · JESHOOTS.com, CC0




