Create spatial environments from a single image
You can now turn a single image into an explorable 3D environment inside Fenestra.
Shaun McCallum
January 19, 2026Late last year, Fenestra was one of the first platforms to support World Labs’ original Marble release.
Today, we’re opening up Marble Mini and Marble Plus to all users - bringing best-in-class spatial generation directly into Fenestra.
These models turn simple image inputs into explorable 3D environments. You can move through them, inspect materials, and experience how scale, light, and form work together in space.
Both models are available from the Spatial section in Fenestra:
Marble Mini
A smaller, contained spatial scene. Ideal for quick tests, experimentation, and understanding how image-to-space works.
(~30 seconds to generate)
Marble Plus
A larger, more detailed environment with richer material depth and stronger spatial presence. This is where the detail really comes through.
(5–10 minutes to generate)
Mini is perfect for fast spatial exploration.
Plus is where the quality shines - sharper, more detail, and far closer to what real spatial presentation can look like.
Compared to many existing Gaussian Splatting solutions, Marble Plus fills key gaps in depth, clarity, and material fidelity.
For best results, we recommend combining it with Nano Banana Pro to:
- Edit spaces
- Change furniture
- Explore design variations
- Generate Worlds
Try it yourself in Fenestra.
Create Spaces In Seconds
Imagine Worlds!Related Articles
Fenestra 101: Sketch to Render, Virtual Staging and AI Editing for Architects & Interior Designers
In this introductory Fenestra webinar, we walk through the core tools architects and designers can use to turn early ideas into visual concepts. The session covers sketch to render workflows, virtual staging, adding furniture, editing images, removing elements and using AI to quickly explore design directions.
Fenestra 101: Our First Webinar
We're running our first ever live webinar on June 3rd. A hands-on intro to Fenestra for architects and interior designers, covering image generation from prompts, 2D and 3D inputs, video, and where AI actually fits in your practice. 30 seats only.
Is conversation the next 3D modelling tool? We tested AI modelling in Rhino
We tested AI 3D modelling in Rhino, driving geometry by conversation, then rendering in Fenestra. Where text-to-3D works, and where it doesn't yet.
