Google began rolling Genie 4 into Project Genie on November 18, 2026, with a rollout for eligible Google AI Ultra subscribers. Google AI Pro members may join a waitlist. Access also requires a Google Account and users must be at least 18; capacity can affect when someone gets in.
Who can access Genie 4 in Project Genie?
Google’s rollout is for eligible Google AI Ultra subscribers, in countries where the plan is available. Google AI Pro members can join a waitlist, and access is granted gradually as capacity allows. High demand may mean a wait even for eligible users; daily usage limits apply and can vary by plan and demand, according to Google’s Project Genie help page.
To use Project Genie, sign in with a Google Account and be at least 18. Those requirements, the plan’s country availability and capacity are separate parts of the access picture: having a qualifying plan does not guarantee immediate entry.
A subscription can make you eligible for the rollout, but age, account access and available capacity still shape when you can use Project Genie.
What Genie 4 adds to Project Genie
Genie 4 lets users change a world with natural-language prompts while exploring it. Google’s examples include changing the weather, adding a character or redirecting the action. Native audio adapts to the unfolding scene, including ambient sound and character dialogue, according to Google’s announcement.
Project Genie also has a conversational agent that can turn a description into initial visuals and a soundtrack. During exploration, Gemini models can offer contextual suggestions or trigger events. A gallery brings together curated and community-created worlds for users to explore, remix and publish.
Genie 3’s generation time and Google’s research challenges
The up-to-three-minute generation length belongs to Genie 3: Google says its limit increased from 60 seconds, and Genie 3 remains selectable in Project Genie. Google identifies camera latency, complex physical interactions, real-time audio precision and long-term scene permanence as continuing research challenges for Genie 3 and Genie 4.
Google also says Project Genie-generated video and audio include SynthID watermarking, and that model inputs and outputs undergo safety screening.