> For the complete documentation index, see [llms.txt](https://docs.altura.com/altura-documentation/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.altura.com/altura-documentation/altura-intelligence/world-engine.md).

# World Engine

## Overview

\
The Worlds Engine leverages generative AI models tailored for 3D asset creation and procedural world-building within Unity and Unreal Engine. It generates terrains, structures, foliage, and environmental details while ensuring technical feasibility and aesthetic consistency.

## Model Type & Architecture

* Multimodal Generative Models: The Worlds Engine uses a combination of 3D generative diffusion models and transformer-based architectures trained on large datasets of game environments. These models generate heightmaps, textures, meshes, and environmental objects that seamlessly integrate into Unity or Unreal.
* Procedural Generation + LLM Integration: LLMs guide high-level thematic and narrative consistency. For example, a text prompt describing a “mystical forest” informs the generative model, which produces terrain, trees, and ruins aligned with that theme.
* Reinforcement Learning for Aesthetic & Performance Optimization: RL is used to fine-tune world generation parameters, striking a balance between visual quality, game performance, and player navigation. The model iteratively refines generated worlds based on developer feedback and performance constraints.

## Underlying Tech

* 3D Diffusion Models: State-of-the-art diffusion architectures adapted for 3D asset generation.
* Text-to-World Pipelines: LLM-based prompt interpretation, guiding generative models with thematic constraints.
* Integration with Unity/Unreal APIs: Specialized adapters ensure the generated assets are compatible, with automated optimization passes for polygon counts, texture sizes, and physics properties.