Purpose/Significance The scoping review summarizes the application of large language models in clinical practice, and provides references for their promotion. Method/Process PubMed, Embase, Wanfang and CNKI databases are searched to screen the literature related to the application of large language models in clinical practice, and the content of the included literature is extracted, summarized and analyzed. Result/Conclusion Large language models have application value in providing treatment suggestions, assisting disease diagnosis, health education, analyzing text image data, etc. However, their performance in answer accuracy and individualization is not satisfactory. In general, large language models show significant potential in clinical practice, but necessary measures must be taken to control the application risks and confirm the scope of application.