This paper presents a preliminary scoping review of research on the application of large language models (LLMs) in video games, analyzing 76 papers published between 2020 and 2024. The review focuses on four main areas: game AI and agents, game development and play, narrative, story, and dialogue, and game research and reviews. The majority of the papers (35.5%) focused on game AI and agents, with LLMs being used for tasks such as reasoning, playing games, and simulating emotions. A significant portion (32.9%) of the papers addressed game development and play, exploring areas such as game level generation, game design, and serious games. Narrative and dialogue generation accounted for 22.4% of the papers, while game research and reviews made up 9.2%. The study found that GPT was the most commonly used LLM, followed by LLaMA, Codex, and BERT. The research highlights the potential of LLMs in enhancing game development, AI, and narrative, but also notes challenges such as lack of logical reasoning and unpredictability. The review aims to provide a snapshot of the current state of research and to lay the groundwork for future studies in this area.This paper presents a preliminary scoping review of research on the application of large language models (LLMs) in video games, analyzing 76 papers published between 2020 and 2024. The review focuses on four main areas: game AI and agents, game development and play, narrative, story, and dialogue, and game research and reviews. The majority of the papers (35.5%) focused on game AI and agents, with LLMs being used for tasks such as reasoning, playing games, and simulating emotions. A significant portion (32.9%) of the papers addressed game development and play, exploring areas such as game level generation, game design, and serious games. Narrative and dialogue generation accounted for 22.4% of the papers, while game research and reviews made up 9.2%. The study found that GPT was the most commonly used LLM, followed by LLaMA, Codex, and BERT. The research highlights the potential of LLMs in enhancing game development, AI, and narrative, but also notes challenges such as lack of logical reasoning and unpredictability. The review aims to provide a snapshot of the current state of research and to lay the groundwork for future studies in this area.