Unity AI Open Beta: Create props with the 3D Object Generator

Unity has opened its AI beta to a new 3D Object Generator that creates static mesh prefabs directly in the Editor. The workflow is built for quick prototyping: describe a prop, optionally supply a reference image, and get a placeholder asset in seconds instead of blocking on concept, modeling, and optimization.

The generator is best suited to simple, single-part objects such as household items, set dressing, and other environment props that don’t need independent animation or complex behavior. Output lands as a standard Unity prefab with mesh and materials, so it can be dropped into scenes and later replaced with final art without disturbing layout or references.

There are two entry points. Developers can use the Generators window for more control over model selection and image inputs, or ask the AI Assistant in natural language and let it handle the full pipeline. If no image is provided, the Assistant first creates a reference image and then generates the 3D object; if an image is included, it uses that directly and can preprocess it as needed.

Unity says clearer, isolated reference images produce more predictable results than text alone. Generated assets are embedded with metadata marking them as AI-generated, which helps with identification and app store disclosure tracking. The beta is available now for Unity 6 developers, and Unity notes the feature set is still under active development and may change.