<p>This study conducts an in-depth study on the automatic generation of infographics by artificial intelligence algorithms. A model based on the Yi 7B large model and integrated with an adaptive chart generation mechanism is constructed to automatically extract features from multimodal data such as text, tables, and images and generate infographics. Through comparative experiments with multiple traditional models such as GraphGen and MultiModalVis, the results show that this model performs well in terms of accuracy, semantic relevance, and layout rationality of chart generation. However, the running speed is not satisfactory when facing large-scale data. It shows good adaptability in data applications of different data types, scales, and industries. The research has problems such as incomplete data set coverage, limited ability to process fuzzy user feedback, and high computing resource requirements. In addition, the current model has weak adaptability to some extreme data scenarios. This study provides a new and effective method for the automatic generation of infographics, and also clarifies the direction of improvement for subsequent research.</p>