Skip to main content
Virtual staging professionals typically spend anywhere from 30-60 minutes lifting a single image of a room into a 3D scene. This process usually involves estimating the camera parameters (e.g. Fspy), modeling surfaces in the image (e.g. 3ds Max), and more. Fill 3D completely automates this process in just 30-50 seconds.

​
Creating a Blender Scene

In the chat interface, select the blend-1 AI model then upload your image: creating a scene In under a minute, the model will respond with a Blender 3D scene you can download and stage manually: staging the scene

​
Sharing the Scene

As a real estate professional, you will often outsource your virtual staging tasks to a professional editor. You can significantly reduce costs and save time by sharing the chat and having them download and stage the generated Blender scene: sharing a chat

​
Ensuring the Best Results

To guarantee best results, below are some guidelines for your image:
The model requires an empty floor plan to properly infer the boundaries of the room. Ensure that the floor is empty, and that the walls are clearly visible.
We are working on adding support for item removal, which will loosen the requirement for empty floors.
The model works best with rectangular rooms. That said, the model will work with rooms of other geometries to varying degrees.
The model requires images with good lighting, because it will mimic the lighting in the 3D scene.
It is highly recommended to have a leveled camera (e.g. using a tripod) when capturing images so as to maximize the field of view of the image.