Large Language Models and Non-Player Characters in Gaming: A Bibliometric Overview

Abstract

This study provides a bibliometric overview of research on large language models (LLMs) in gaming, with a particular focus on their application to non-player characters (NPCs). The analysis covers 35 documents published between 2022 and 2025 and reveals a sharp increase in publication volume and citations, indicating growing scholarly interest. The results indicate that key publication venues include arXiv and IEEE Transactions on Games, with Microsoft Research emerging as a leading contributor. Overall, the findings reflect an interdisciplinary and rapidly evolving research landscape, centered on leveraging LLMs to develop intelligent, interactive NPCs and enhance game design through dynamic player-NPC interactions.