TableGPT-agent is a pre-built agent model based on TableGPT2, designed for question-answering tasks involving tabular data. Developed using the Langgraph library, it offers a user-friendly interface and efficiently handles complex table-related questions. TableGPT2 is a large multimodal model that combines tabular data with natural language processing, providing powerful support for data analysis and knowledge extraction. This model is suitable for scenarios requiring fast and accurate processing of tabular data, such as data analysis, business intelligence, and academic research.