Code2Games: Enabling Coding Agents for Gaming World Generation

2026-10-04

AI Summary

This study addresses the insufficient consistency in scenes, layouts, and gameplay logic when existing coding agents generate game worlds by proposing the Code2Games framework. This method coordinates scene analysis and gameplay planning through shared representations, introducing a structured Blender foundation alongside a persistent element correspondence mechanism. By integrating multi-agent collaboration, constraint-based generation, and Unreal Engine 5 adaptation, it leverages compilation diagnostics and runtime feedback to guide execution refactoring, thereby eliminating inconsistencies. Furthermore, this work introduces GameCode4D, a pioneering benchmark enabling systematic multi-dimensional evaluation. Experimental results demonstrate that the proposed approach significantly enhances the visual quality, interactive fidelity, playability, and overall quality of multimodal artifacts within the generated game worlds.

Abstract

Generating a high-quality gaming world from a natural-language game intent requires joint reasoning about scene structure, spatial layout, gameplay objectives, interactive entities, and executable gameplay logic. Existing coding agents can generate individual assets, scenes, or scripts, but often struggle to maintain consistency across these components. We propose Code2Games, an agentic framework that builds a structured gaming world upon a base Blender world generated from the same game intent. Code2Games coordinates scene analysis, gameplay planning, constrained gaming-world generation, and gaming-engine customization through a shared scene-gameplay representation with persistent element correspondence. After world generation, Code2Games adapts the generated world to Unreal Engine 5 and employs an execution-guided reconstruction process that uses compilation diagnostics, runtime feedback, and gameplay test results to resolve inconsistencies arising during engine adaptation. To systematically evaluate gaming-world generation, we introduce the GameCode4D benchmark, which comprises ten fixed game prompts spanning different levels of scene and gameplay complexity. We evaluate the generated results across four dimensions: visual quality, interactive fidelity, multimodal artifact quality, and playable-game quality. Experiments demonstrate that, compared with direct gaming-world generation by coding agents and existing baseline methods, Code2Games consistently improves the visual quality and interactive fidelity of generated gaming worlds, as well as the quality of the resulting games after engine adaptation.

Problem

• Gaming World Generation
• Coding Agents
• Cross-component Consistency
• Engine Adaptation
• Evaluation Benchmark

Innovation

• Coding Agents
• Gaming World Generation
• Agentic Framework
• Execution-Guided Reconstruction
• GameCode4D Benchmark