Google 预览 Genie 4 世界模型，可生成带音频的 3D 环境

Google LLC today previewed Genie 4, an artificial intelligence model that can generate realistic three-dimensional environments.

The algorithm is the latest addition to a series of world models that the company launched in 2024. Compared with its predecessor, Genie 4 is better at rendering fluid movements and other complex details. Furthermore, it can generate accompanying audio such as the sound of crashing waves.

“The most dramatic leap is how Genie 4 simulates people, creating characters that move naturally, respond organically to their environments, and can be interactively steered in real-time,” Google DeepMind staffers Raia Hadsell, Diego Rivas and Hang Qi wrote in a blog post.

Users can create a virtual world by providing Genie 4 with a natural language description of what they wish to render. They can then navigate the simulation and make edits by entering follow-up prompts. For example, a user could ask Genie 4 to change the weather or add more objects. The model can make such edits in a manner that doesn’t reduce the virtual world’s photorealism or consistency.

Genie 4’s natural language interface is available alongside keyboard controls. They can be used to change the position and angle of the camera.

Under the hood, Genie 4 uses several other Google models to boost its output quality. It relies on Gemini Omni, a video generator the search giant introduced in May, to ensure that the visual assets it generates are realistic. Gemini 3.8 Flash, a reasoning model optimized for cost-efficiency, makes the individual frames of a simulation consistent with one another.

“Gemini continuously observes the scene, breaks down goals into a sequence of steps, and reviews incoming frames to track the progress and verify that each step succeeded,” Hadsell, Rivas and Qi wrote. “As the world shifts or the user intervenes, Gemini adapts its plan in the moment.”

DeepMind is currently studying how Genie 4 can be applied to robotics use cases. The project combines the algorithm with Gemini 4 Argon, Google’s flagship reasoning-optimized large language model.

The project’s goal is to let engineers simulate robots using high-level instructions. For example, a user could ask Argon to generate synthetic footage of a robotic arm soldering together two car parts. Argon can turn the brief request into dozens of highly detailed, multi-sentence Genie 4 prompts that each describe a different step of the soldering workflow.

Increasing the number and precision of the prompts provided to Genie 4 improves rendering quality. That quality boost, in turn, makes the model’s robot simulations more useful for AI training purposes. Developers train AI models to control robots by providing them with visual examples of how to complete the tasks they will be expected to perform.

Google is currently testing Genie 4 with a limited number of early adopters. It plans to make the model available via its Google AI Ultra subscription for consumers, the Gemini API and Google AI Studio.

Image: Google