Text to 3D

Generate a textured 3D model from a written description

Text to 3D builds a model from your prompt alone—no reference image needed. Describe the object, pick a model that accepts text input, and the node returns a textured 3D asset you can stage, capture, and export.

How It Works

1. Add a 3D Node and pick a model with text input (see 3D Models).

2. Type a description directly into the node, or connect a text node to its prompt input.

3. Click Run. Most generations take around 5 minutes.

Some models (for example Tripo P2 and Tripo H3.1) also accept a negative prompt and per-stage seeds so you can steer away from unwanted details and reproduce a result.

Prompting Tips

• Describe one object — text-to-3D models produce a single asset; scenes with several subjects tend to merge or drop elements

• Name material and style — "matte ceramic", "brushed steel", "low-poly cartoon" translate directly into textures and geometry

• Keep the pose neutral — front-facing, symmetric descriptions reconstruct more cleanly than dynamic action poses

• Iterate on 2D first — if a look needs several passes, refine it with an image model and switch to Image to 3D

Models

Not every 3D model supports text input; the model picker only offers compatible models once a prompt is the sole input. Visit 3D Models for the full list with descriptions and supported inputs.