Skip to content
NewA calmer, faster ZeroHour
StoryModel releaseinfo

DeepMind previews Genie 4 world model with native audio

6 sources · 6 articlesFirst reported Updated ()1
What's new

October 28: A MarkTechPost item headlines Genie 4 as a real-time model with native audio, but its page body is only a cookie notice and adds no new facts.

Merged summary · x-ai/grok-4.7 · rewritten as coverage arrivesSummaries by Vani AI

Genie 4 is a limited research preview that generates short interactive worlds with native audio, with wider access planned next.

Google DeepMind announced Genie 4, a research-preview world model that generates short interactive simulations with native ambient sound and dialogue. It improves on Genie 3 with more realistic physics and people; Gemini 3.8 Flash plans steps and checks frames so users can change weather, objects, people, or the camera in real time, while Gemini Omni anchors appearance and behavior. Google says scenes stay consistent for about two minutes, double Genie 3, but details can drift and audio accuracy and in-world text remain weak. Access starts with trusted testers; a Project Genie waitlist is open, and Google AI Ultra subscribers and early Gemini API partners follow. DeepMind is testing Gemini 4 Argon with Genie 4 on robotics, including a simulated Apptronik Apollo humanoid playing checkers, and says generated video and audio will carry SynthID.

Key facts
  • Trusted testers have Genie 4 now; Ultra and early Gemini API access follow.
  • Scene consistency is about two minutes, double Genie 3’s one-minute limit.
  • Gemini 3.8 Flash plans steps and checks frames during real-time edits.
  • Gemini Omni anchors appearance and behavior; Gemini 4 Argon is in robotics tests.
  • Generated video and audio will carry Google’s SynthID watermark.
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

What each outlet adds

beyond the shared facts
  • The VergeCites film, game, and robot uses and an open Project Genie waitlist.
  • Google DeepMindNames Gemini Omni and calls Genie 4 its first native-audio real-time world model.
  • TechCrunchSays audio accuracy and legible in-world text remain unsolved.
  • The DecoderAdds SynthID watermarking and an Apollo checkers robotics demo.
  • Ars TechnicaFrames paid access as the likely near-term route to trying Genie 4.
Vendors
Google DeepMindGoogle
Products
Project GenieGemini APIGoogle AI StudioSynthIDGeminiVeo
Organizations
Google DeepMindApptronik
AI models
Genie 4Gemini OmniGemini 3.8 FlashGemini 4 ArgonGenie 3

Coverage

oldest first · one row per article · UTC
  1. · 3d ago The Verge · AIFirst report

    Google announces Genie 4 for longer, more realistic AI simulations

    Experiences last about two minutes, up from one minute in Genie 3, and a Project Genie waitlist is open.

  2. · 3d ago Ars Technica · AI

    DeepMind Genie 4 world model adds audio and more realistic people

    Gemini 3.8 Flash guides each frame, but DeepMind says simulated worlds stay consistent for only a couple of minutes.

  3. · 3d ago Google DeepMind

    DeepMind announces Genie 4 real-time world model with native audio

    A research preview goes to trusted testers first, then Google AI Ultra subscribers in Project Genie and early Gemini API partners.

  4. · 3d ago TechCrunch · AI

    Google DeepMind releases Genie 4 world model in research preview

    Trusted testers get it now; Google AI Ultra and Gemini API partners follow in the coming weeks.

  5. · 3d ago The Decoder

    Google DeepMind unveils Genie 4 real-time world model as research preview

    Only select testers have it now; Google AI Ultra subscribers and early Gemini API partners are due next, with SynthID on outputs.

  6. · 2d ago MarkTechPost

    Google DeepMind introduces Genie 4 real-time world model

    The title says Genie 4 includes native audio, but the captured page is only a cookie notice.