Google DeepMind unveils Genie 4, a natively multimodal world model with real-time audio
scaling01 · x · 2026-10-28
Google DeepMind announced Genie 4, a new world model with unprecedented realism and real-time interactive simulation. It's natively multimodal — generating music-like audio, ambient sounds, and interactive dialogue alongside the world itself. Built on the Gemini family (Gemini Omni grounds visuals; other Gemini models provide reasoning for coherence), Genie 4 simulates what happens next after you set an event in motion. DeepMind frames world simulation as crucial to AGI, targeting robot/agent training environments and new interactive experiences. Rolling out as a research preview to trusted testers, then Google AI Ultra subscribers via Project Genie and Gemini API/AI Studio partners.
Related event: DeepMind Unveils Genie 4, Its Most Interactive World Model(17 posts)→
More from Embodied
- NVIDIA Launches SimReady Central: 3,500+ Free OpenUSD Simulation Assets on Hugging Face — mishig25 · 2026-10-28
- Formic CEO Co-Writes Essay on the Realistic Path to General-Purpose Robots That Don't Suck — AnneliesGamble · 2026-10-28
- YC: As robot skills transfer across machines, the bottleneck shifts to hardware — ycombinator · 2026-10-28
- Xplore launches RWE-Bench: physics-verified RL training tasks for engineering — Jsevillamol · 2026-10-28
- Gemini watches Genie 4 render frame-by-frame to simulate Apollo humanoid playing checkers — chrisfirst · 2026-10-28
- Robot learns new task from a single demo: Machenta unveils Vision-Demo-Action foundation model — Ed__Johns · 2026-10-28