On this page
Google DeepMind’s Genie product page describes Genie 4 as “our most realistic and interactive world model yet.” The same day, @GoogleDeepMind posted on X (2115114028338962672 / 2115114030503166385; created_at 2026-10-27T16:10:55Z (2026-10-28 00:10 Asia/Shanghai)). Related reading: How to Choose Gemini Video Mode—Static 1 FPS or Agentic Retrieval.
Natural-language scene control and native audio
DeepMind says natural language can create and change worlds live: describe landscapes and events, move the camera, add new characters, or change the weather, then see what happens next. Worlds also include native audio—footsteps, crowds, and characters that can speak with players and each other, according to the product page.
Continuous interaction: a few minutes
Under Limitations, DeepMind states that the model can support a few minutes of continuous interaction, rather than extended sessions. That is DeepMind’s published duration wording. For reference only.
Availability: trusted testers first; Ultra / Project Genie coming soon
Genie 4 is available as a research preview for trusted testers, and will be coming soon starting with Google AI Ultra subscribers in Project Genie and early access partners of the Gemini API and Google AI Studio. Related reading: What Is Runway App Studio? How It Differs from Apps and Workflows, Plus Credits and Sharing Limits.
Technical notes
DeepMind says Genie 4 incorporates capabilities from the Gemini family, with Gemini Omni anchoring simulations in real-world appearance and behavior, and other Gemini models supplying reasoning for coherence. Generated video and audio include SynthID watermarking. Media coverage such as The Verge (datePublished 2026-10-27T16:00:00+00:00) may restate the launch; figures here follow DeepMind’s product page and the official X posts. Nearby creator-video coverage: How to Direct Characters, Camera and Pacing in Kling 4.0.
Product details can change; DeepMind’s Genie page carries the latest wording. For reference only.