GenLARP: Enabling Immersive Live Action Role-Play through LLM-Generated Worlds and Characters

TL;DR

GenLARP tackles the accessibility and engagement barriers of traditional LARP by enabling a single user to author and participate in immersive stories in VR using AI-driven agents. The approach combines Narrative Initialization, Interactive Role Design, and Live-Action Role Play within a Unity-based pipeline powered by GPT-4o and SynCity to generate structured narratives, character behaviors, and adaptive 3D scenes. Key contributions include a unified narrative semantic pipeline, stateful, multi-character agents, and a real-time, first-person LARP interface, plus support for non-linear narrative branching and dynamic pacing. The work demonstrates feasibility of AI-assisted, immersive storytelling in VR and outlines concrete paths toward collaborative co-creation, memory-aware coherence, and multimodal scene construction.

Abstract

We introduce GenLARP, a virtual reality (VR) system that transforms personalized stories into immersive live action role-playing (LARP) experiences. GenLARP enables users to act as both creators and players, allowing them to design characters based on their descriptions and live in the story world. Generative AI and agents powered by Large Language Models (LLMs) enrich these experiences.