Skip to content
NewA calmer, faster ZeroHour

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

Original title: Google Deepmind's Genie 4 puts a modern spin on agent designs that predate LLMs

The Decoderby Maximilian Schreiner ()2 min read1
Part of a story covered by 6 sourcesDeepMind previews Genie 4 world model with native audioMerged summary & timeline
AI summary · x-ai/grok-4.7Summaries by Vani AI

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

Google DeepMind introduced Genie 4, successor to Genie 3, as an experimental world model that renders interactive scenes frame by frame, including sound and dialogue, from keyboard input and text prompts. Gemini models, including Gemini Omni and Gemini 3.8 Flash, plan steps in the background so the scene stays coherent. DeepMind says short sessions hold together, but longer runs can shift visuals and drop objects. Access is limited to select testers, with Project Genie planned for Google AI Ultra subscribers and early Gemini API partners.

Unconfirmed: Capability claims, including real-time dialogue and natural character motion, are DeepMind's and the model is described as early and experimental.

Key facts
  • Genie 4 succeeds Genie 3, which DeepMind showed in August 2025.
  • Gemini 3.8 Flash watches the scene and checks whether each planned step worked.
  • A demo used Gemini 4 Argon text to simulate Apptronik's Apollo playing checkers.
  • DeepMind says it is still researching complex physics, accurate sound, and readable text.
  • Generated videos and audio tracks 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

From the original article

2 min read at the-decoder.com

Google Deepmind has introduced Genie 4, a real-time world model that simulates interactive 3D worlds with sound and talking characters. Gemini models help plan in the background to keep the simulation consistent. Genie 4 is still an early research model, and longer sessions lead to visual inconsistencies.

The Decoder

By Maximilian Schreiner · the-decoder.com

Read the full article at The Decoder

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.