Can anyone use Genie 4 now?

No. As of October 28, 2026, Genie 4 is a research preview for trusted testers. Google has not announced a general availability date.

What can Genie 4 do that earlier world models could not?

Genie 4 generates ambient sound, music and conversation alongside visuals, and it responds to movement, viewpoint changes and text instructions. It also integrates with Gemini models for scene observation and progression.

How does Genie 4 differ from Marble or Runway?

Marble offers export formats like Gaussian splats and meshes for production pipelines, while Genie 4's announcement did not confirm similar exports. Runway's GWM Worlds 2 and Odyssey's Starchild-1 are earlier examples of sound-enabled world generation.

Large Language ModelsVideo GenerationAudio & Speechnote AI21h agoNov 2, 2026, 13:00 JST

Google DeepMind's Genie 4 builds walkable, sounding worlds

Send on LINE

3 Key Points

What happened
Google DeepMind announced Genie 4 on October 27, 2026 — a world model that builds scenes you can walk through and change with text, generating visuals, ambient sound, music and conversation.
Why it matters
It moves world models from passive video generation toward interactive, sound-rich environments, but it is a research preview for trusted testers, not a general release.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Get these free every morning

Context & Analysis

Genie 4 arrives after other sound-capable world models: Odyssey's Starchild-1 was announced in May 2026, and Runway's GWM Worlds 2 in September. Google DeepMind's version stands out for its integration with Gemini models—Gemini Omni supports appearance and behavior, while Gemini 3.8 Flash observes scenes and adjusts procedures—and for combining visuals, ambient sound, music and conversation in one interactive experience.

The announcement highlights limitations: long-duration consistency, object permanence, re-creating real locations, and accuracy of text and sound are listed as challenges, and continuous operation lasts only a few minutes. Google also notes that realistic-looking video cannot guarantee design values like dimensions, strength or temperature, so engineering adoption would require analysis and measurement that the preview does not yet support.

Access is gated. Google points to Project Genie's Google AI Ultra subscribers and early-access partners via Gemini API and Google AI Studio as future expansion paths, but no Japan general availability date, Genie 4-specific pricing, or free tier was disclosed as of October 28, 2026. For now, the practical step is to check the official announcement's waitlist and review the terms before applying.

FAQ

Can anyone use Genie 4 now?: No. As of October 28, 2026, Genie 4 is a research preview for trusted testers. Google has not announced a general availability date.
What can Genie 4 do that earlier world models could not?: Genie 4 generates ambient sound, music and conversation alongside visuals, and it responds to movement, viewpoint changes and text instructions. It also integrates with Gemini models for scene observation and progression.
How does Genie 4 differ from Marble or Runway?: Marble offers export formats like Gaussian splats and meshes for production pipelines, while Genie 4's announcement did not confirm similar exports. Runway's GWM Worlds 2 and Odyssey's Starchild-1 are earlier examples of sound-enabled world generation.

note AIRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.