Domain models in Artificial Intelligence (AI) Planning represent the essential knowledge required for planning tasks in various sectors such as smart buildings, robotics, and autonomous vehicles. These models are critical for defining the actions, conditions, and effects that an AI planner uses to generate plans. However, due to their complex and hierarchical nature, creating these domain models is a complex and time-consuming process, especially for Hierarchical Task Network (HTN) based models. Recognizing this challenge, this research investigates the use of Large Language Models (LLMs) to automatically generate domain models for AI planners. LLMs have previously demonstrated impressive abilities in generating code and other specifications from natural language inputs. While there has been preliminary research on using LLMs to produce non-hierarchical domain models specified in the Planning Domain Definition Language (PDDL), there has been no study focused on generating HTN domain models. This thesis proposes a structured prompt engineering approach to evaluate the effectiveness of LLMs in generating domain models in Hierarchical Planning Definition Language (HPDL). We conducted experiments using several domain datasets, including cafeteria, blocks world, satellite, and a general domain, employing advanced models such as Generative Pre-Trained Transformer (GPT)-3.5, GPT-4, and GPT-4o. An SH planning system was utilized to verify the generated models. The results indicate that LLMs can proficiently translate natural language into structured planning languages. However, challenges were encountered in generating plans that require arithmetic and physical reasoning. Despite these difficulties, the structured prompt technique significantly enhanced the potential of HPDL models, reducing the need for extensive human intervention. These findings suggest that LLMs can effectively bridge the gap between planners and human users, laying a foundation for future research in their use for creating advanced and accurate domain models. This research highlights the potential of LLMs as efficient tools for automated domain model generation in AI planning.