Abstract: With the aggravation of global environmental issue,the low-carbon electricity market has garnered significant attention. This article delves into the current development status and the challenges faced by the low-carbon electricity market,with a special focus on the potential applications of large language models（LLMs）in advancing the development of this market. The paper emphasizes that through fine-tuning,prompt engineering,and semantic embedding,LLMs can effectively adapt to specialized applications in the power sector,particularly showing promise in areas such as adjusting power source structures,predicting electricity demand,and risk alerting.Furthermore,methods based on LLMs,including agents and the chain of thought approach,can solve complex problems and contribute to the construction of the low-carbon electricity market. With the development of the technologies behind LLMs and the deepening of power industry reform,these methods are anticipated to facilitate the low-carbon transition of China’s power system and contribute to the realization of its dual carbon objectives. Nevertheless,the inherent limitations of LLMs must be recognized,and proactive steps should be undertaken to safeguard against potential risks,ensuring their secure and orderly development in the low-carbon market.