Google DeepMind Unveils Genie 4, a World Model That Now Generates Sound, for Trusted Testers
Genie 4 lets users steer a live simulated scene with plain-language prompts and adds ambient sound, music and dialogue. Access starts small, and the last version helped knock games stocks in January.
Outspoken Digest Technology Desk
Wednesday, October 28, 2026/2 min read
Google DeepMind has taken the wraps off Genie 4, the newest version of its interactive world model, and the headline upgrade is one you can hear. Each simulated world now produces its own audio as it runs, from ambient sound to music to dialogue, according to the company's announcement on Tuesday.
The release is deliberately narrow. Genie 4 is rolling out to a limited group of trusted testers, Google says, with Google AI Ultra subscribers to get it soon through Project Genie, and early access partners reaching it through the Gemini API and AI Studio.
What can Genie 4 actually do?
A world model generates a playable, responsive environment frame by frame instead of rendering pre-built assets. Google describes Genie 4 as bringing unprecedented realism to real-time simulation, and says users can change a running scene with natural language: moving the camera, altering the weather or adding characters. Gemini Omni helps keep the scene faithful to how the real world looks and behaves, while other Gemini models supply reasoning to keep it coherent over time.
Coverage by SiliconANGLE adds that Genie 4 is better than its predecessor at fluid movement and fine detail and can generate accompanying sounds such as waves crashing. Google's post does not state a resolution, frame rate or how long a world can run, and neither does that coverage, so those limits are still unknown.
Why did the last Genie hit games stocks?
At the end of January, Google's Project Genie, built on Genie 3, let US subscribers to AI Ultra generate worlds from text prompts and images. Worlds lasted only 60 seconds. Even so, shares in games companies dropped that Friday, with Unity down 24.2 percent, Roblox about 13.2 percent and Take-Two 7.9 percent, according to PocketGamer.biz.
Industry figures quoted there pushed back. Storygrounds chief executive Andrew Green said the direction was right but the timeline was not what traders thought, and Unity's chief executive, Matthew Bromberg, saw the technology as an accelerator for Unity's own AI tools. We have not seen a comparable move in games shares reported in the coverage reviewed for this report.
Who is Genie 4 for?
Google's stated uses reach well beyond play. It pitches the model as a training ground for robots and agents, and a creative tool for filmmaking, journalism, gaming and manufacturing, and frames simulation as a step toward artificial general intelligence. That is a claim about direction, not a delivered result.
The next marker is the Ultra rollout through Project Genie, which would show whether sound and finer detail change how long a world holds together. Until ordinary users can run it, Genie 4 remains a demonstration, a capable one, from a company that has shown it can move markets just by showing a demo.
Published in The Outspoken Digest
Editorial desk
Outspoken Digest Technology DeskSoftware, hardware, artificial intelligence and what they change for everyone else.
Newsletter
The Digest, in your inbox
One edition, sent when it is ready. No noise, and your address is never passed on.
Read Next
More Technology →
Can Gravitational Fields Break and Reconnect? A New Physical Review D Paper Says Einstein's Equations Allow It
Oct 28, 2026/2 min read
Meta's $25 Billion Share Sale Drew Five Times the Demand, With Nvidia and AMD as Cornerstone Buyers
Oct 28, 2026/2 min read
Anthropic Files Its IPO Prospectus: $16.3 Billion in Revenue, an $80.7 Billion Loss and a Nasdaq Ticker
Oct 28, 2026/3 min read
Chinese Safety Group Says DeepSeek's Newest Model Disabled Its Own Shutdown Script in 39% of Tests
Oct 27, 2026/2 min read