## The Modelling Barrier
For decades, three-dimensional spatial design required fluency in complex software — Rhino, Revit, SketchUp, 3ds Max. These tools are powerful, but they demand significant investment in time and training. The result: a gap between the designer who has the spatial idea and the modeller who can realise it.
Text to 3D closes that gap entirely.
## How It Works
OctopusWave's Text to 3D system converts natural language descriptions into navigable GLB models — the open standard format compatible with all major 3D viewers, game engines, and AR platforms.
The process:
1. Add a Text to 3D node to your canvas
2. Describe the space you want to create
3. Generate — the system returns a navigable 3D model within the node
The model is immediately interactive: you can orbit, zoom, pan, and inspect it from any angle directly inside OctopusWave. Export to GLB for use in Blender, Unreal Engine, Unity, or any WebGL viewer.
## Describing Space Effectively
The quality of the output depends on the specificity of the description. Vague prompts produce generic results; architectural descriptions produce architectural output.
**Example prompt:** *A double-height entrance hall with exposed concrete walls, a cantilevered steel staircase rising from left to right, floor-to-ceiling glazing on the north facade, and polished concrete flooring reflecting soft diffused daylight.*
The system understands spatial relationships (double-height, cantilevered, north-facing), material behaviour (exposed concrete, polished reflection), and light quality (diffused, floor-to-ceiling glazing) — because it was trained on architectural language.
## Use Cases
**Concept exploration:** Generate multiple spatial configurations from text descriptions before committing to any one scheme. Compare a linear plan against a courtyard typology in minutes.
**Client communication:** Give clients a navigable model before detailed drawings are complete. Spatial experience is more persuasive than plan diagrams.
**Design development:** Use the 3D output as a starting point for further modelling in dedicated software — shortcutting the early massing stages.
## Getting Started
Text to 3D is available from the Starter plan onwards. Add a 3D node to any canvas from the Quick Add menu, describe your space, and generate.