Google opens Genie 4 to AI Ultra subscribers
Original title: Genie 4 is now available
Eligible subscribers can edit worlds mid-session with native audio, while Genie 3 runs extend to three minutes.
Google DeepMind and Google Labs rolled out a new Project Genie experience to eligible Google AI Ultra subscribers on November 18, 2026. It adds first access to Genie 4, Genie 3 upgrades, a Project Genie agent, and a community gallery for discovering and remixing worlds. Genie 4 supports real-time prompt edits and native audio. Generated video and audio include SynthID watermarks, and inputs and outputs pass safety screenings. Google AI Pro members can join a waitlist.
- Genie 3 sessions can last up to three minutes, up from 60 seconds.
- Users can seed Genie 3 worlds from 360-degree panorama images.
- The agent uses Gemini models for visuals, scores, and in-session suggestions.
- Google says Genie 3 and Genie 4 remain frontier research world models.
- Vendors
- Organizations
- Google DeepMindGoogle Labs
From the original article
4 min read at deepmind.googleExplore a new Project Genie experience for Google AI Ultra subscribers, featuring Genie 4, the Project Genie agent, and a community gallery. This content is generated by Google AI. Generative AI is experimental Earlier this year we introduced Project Genie , an experimental research prototype that lets you generate and explore interactive worlds.
deepmind.google
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.