Large language models, as the core technology of natural language processing, have achieved capabilities in text representation, inference, and generation through massive-scale language pre-training. This has brought broad application prospects in the financial sector. This paper first introduces the development history of large language models, then focuses on their typical applications in financial areas such as risk management, credit rating, fraud detection, algorithmic trading, and financial forecasting. Finally, it briefly discusses the main challenges currently faced by large language models in the financial sector, as well as future research directions. This paper aims to provide a comprehensive understanding of large language model technology and its current applications, challenges, and prospects in the financial sector.