Skip to main content

Prompting for 3D models

Basic important variables​

To achieve the best results and 3D objects that closely match your vision when using Artificial Intelligence, it's crucial to formulate precise and detailed prompts. The more specific information you provide, the better the AI can understand and create what you want.

For optimal results, it's recommended to clearly define the following variables in your prompt:

  1. Type: Clearly name the object you want to generate (e.g., "office chair", "old sports car", "minimalist coffee cup").
  2. Style: Define the aesthetic or graphical style (e.g., "realistic", "low-poly", "science-fiction style", "cartoonish", "cyberpunk").
  3. Material: Specify what the object should be made of (e.g., "wooden", "metallic", "glassy", "plastic with a matte finish", "fabric").
  4. Color: Provide specific colors or a color palette that should dominate (e.g., "red with chrome elements", "pastel color palette", "black and white").
  5. Size: While scale can often be adjusted later, specifying the relative size or proportions in the prompt can be helpful (e.g., "small", "large", "massive", "long and narrow").
  6. Wear: If applicable, describe the desired condition of the object (e.g., "new and shiny", "slightly worn", "old and rusty", "battle-damaged").
  7. Similarity: If you are basing it on a specific idea or reference, you can indicate how similar the generated object should be to it, or describe the key features that should be retained. Remember that Ludus also sees your viewport so it can be based on things already in the scene. It can even be a screenshot.

Precisely defining these parameters significantly increases the chance of obtaining a 3D object consistent with your vision.