Google DeepMind Releases Research Preview of World Model Genie 4

Home | News | Blog | Academicians | Talent | Conferences | Funds & Projects | Papers | Graphics | Video & Live | Xiaoke Robot | Medical Science Popularization

Author: Yu Yan Source: The Paper Published: 2026/10/28 13:19:25

Select font size: Small Medium

Large

Google releases next-generation world model Genie 4, strengthening its Physical AI technology layout

On October 27 local time, Google's DeepMind released a research preview of its next-generation world model Genie 4. The model further enhances real-time interactive world simulation capabilities while exploring applications in robot training and the field of Physical AI.

Compared with the previous-generation model Genie 3, Genie 4 not only improves in visual realism, physics simulation and naturalness of character movements, but also extends the duration of scene-consistent simulation to about two minutes, twice that of the previous model.

According to the introduction, Genie 4 can generate the most realistic, real-time interactive dynamic virtual worlds to date, achieving progress in simulating physical phenomena, character movements and scene interaction. Users can use natural language instructions to change the weather, adjust the viewpoint, or add characters and objects in real time during simulation. The model also has native audio generation capabilities, able to synchronously generate background music, ambient sounds and interactive dialogue.

Notably, Google also demonstrated an experiment in which Genie 4 works with the large model Gemini 4 Argon to simulate a humanoid robot performing tasks.

In the demonstration, Gemini 4 Argon was responsible for breaking down user instructions into multiple steps, while Genie 4, based on these instructions, generated in real time simulated footage of the robot performing the task, showing the Apollo humanoid robot from the US robotics company Apptronik playing checkers.

This experiment also demonstrates Google's technology layout in the field of Physical AI: attempting to combine the task planning and reasoning capabilities of large models with the environment simulation capabilities of world models, enabling AI to simulate different tasks and their execution processes in virtual worlds.

Genie 4 brings more realistic visuals, smoother motion effects, and supports native audio generation. Genie 3 on the left, Genie 4 on the right. Source: Google DeepMind official website

Tejas Kulkarni, Google's director of research, said at a media briefing before the release that more and more "hybrid world models" will emerge in the future, in which Gemini generates code and text, and Genie receives this information in real time and simulates the real world at different levels of abstraction.

World models are becoming an important technology direction for tech companies. Potential applications cover not only game development, interactive education and film and television production, but also robot training, autonomous driving simulation and other fields. By generating interactive virtual environments, world models are expected to help robots and AI agents conduct task simulation and testing before entering real environments.

Industry competition and integration are also accelerating. In September this year, AI video generation company Runway released its second world model, GWM Worlds 2, which supports dynamically generating interactive scenes based on user input and has audio generation capabilities.

On the capital market side, on September 28, AMD announced it would acquire World Labs, founded by Chinese-American AI scientist Fei-Fei Li, in an all-stock transaction worth about $8.2 billion, strengthening its technology layout in world models, spatial intelligence and Physical AI. Unlike Google's Genie 4, which focuses on real-time dynamic simulation, World Labs' previously launched Marble mainly focuses on generating three-dimensional worlds with spatial consistency that can be explored and edited over the long term.

In addition, according to The Wall Street Journal, Amazon is in talks to acquire world model startup Decart for about $7 billion. On October 22, US electric vehicle manufacturer Rivian announced the acquisition of world model startup Third Dimension, planning to apply the related simulation technology to the development and testing of autonomous driving systems.

However, world models still face technical challenges before large-scale practical application. DeepMind engineer Diego Rivas said at a media briefing that Genie 4 is still in an early stage, and the complete product application form is not yet clear. As simulation time lengthens, details and objects in the scene may drift, and issues such as audio accuracy and the clarity of text in the visuals remain to be resolved.

Google plans to open it to Google AI Ultra subscribers through Project Genie in the coming weeks, and gradually provide related capabilities to early access partners of the Gemini API and Google AI Studio. Special statement: This article is reprinted solely for the purpose of disseminating information and does not mean that it represents the views of this website or confirms the authenticity of its content; if other media, websites or individuals reprint and use it from this website, they must retain the "source" indicated by this website and bear their own legal responsibilities such as copyright; if the author does not wish to be reprinted or wants to contact us regarding reprint fees and other matters, please contact us.

Short burst and long burst appear successively 4 minutes apart: new discovery of "double explosion" challenges traditional theory of gamma-ray burst origins

3

Brief Report | Team of Professor Li Min from Shanxi Provincial People's Hospital publishes brief report: Trends in gout-related cardiovascular disease mortality among US adults aged 55 and older, 1999-2024

4

GCB | China Agricultural University reveals the divergent trajectory of available phosphorus in Chinese farmland from deficiency to enrichment over the past 40 years

5

PCE | Ma Chuanxi/Si Hongqi research group of Anhui Agricultural University reveals that the miR156h-TaSPL4–TaPIN18 module regulates wheat plant architecture and grain size through auxin transport

6

Double harvest in plastic reduction and efficiency enhancement: NiCo2O4 electrocatalysis helps resource utilization of PET waste

7

Information journal's 2025 Impact Factor rises to 4.3, an increase of 48%

8

Biology | Highly cited articles in the Plant Science section, 2024-2025

Cover article | Mycology: Unlocking the hidden "colors" of fungi

Self-powered triboelectric electronic skin "reads" silent lip language, opening a new paradigm for human-computer interaction and virtual communication

Defect engineering synergistically regulates oxygen vacancies and domain structure of CaBi4Ti4O15 ceramics