Large language models in artificial intelligence are a natural language processing technology based on deep learning. By training on large amounts of corpus data, they simulate the generation process of human language. Traditional Chinese medicine theory and practice involve a wealth of text data, such as classical works, medical cases, and prescriptions. This text data contains a wealth of medical terminology, disease names, symptom descriptions, and other information, which is a valuable source for natural language processing tasks. Large language models can learn and train on this text data to grasp the knowledge and language rules of traditional Chinese medicine, and thus be applied to natural language processing tasks in the field of traditional Chinese medicine. In addition, large language models have strong semantic understanding and reasoning capabilities, which can help doctors better understand patients' conditions and symptoms and provide more accurate diagnosis and treatment plans. Therefore, the application prospects of large language models in traditional Chinese medicine are very broad. This article reviews the current status of large language models in disease diagnosis, herbal medicine research, and medical knowledge base construction in traditional Chinese medicine, discusses the existing problems and future research trends, and aims to better think about the direction and future development of intelligent traditional Chinese medicine by clarifying the difficulties and bottlenecks faced by large language models in the field of traditional Chinese medicine research.