EXTENDED ABSTRACT: Large models, as a core component of artificial intelligence, have constructed complex and powerful data processing systems through core technologies such as deep learning, pre-training, and transfer learning. They not only efficiently process massive amounts of data, demonstrating excellent generalization capabilities, but also drive the widespread application of artiffcial intelligence in scientiffc ffelds. This report outlines the progress made by Shanghai Artiffcial Intelligence Laboratory in recent years in empowering chemical and materials science research with large models, covering a full chain of achievements from computational simulations, scientiffc language models to intelligent experimental result analysis. Machine learning potential functions, which use machine learning techniques to simulate intermolecular forces and system energies, can efficiently predict material properties and accelerate the development of new materials, gradually showing their great potential in the ffelds of chemistry and materials science. Chemical large language models can predict and generate molecular properties, design chemical experiment protocols, and enhance their professionalism in the ffeld of chemistry through integration with external tools and databases. Spectroscopic large models, serving as the "eyes" for chemical and materials research, play a crucial role in material structure elucidation, property prediction, and reaction mechanism understanding. The application of large model technology in chemical and materials science research is of great signiffcance, not only deeply empowering intelligent regulation and precise generation in chemical research, but also accelerating breakthroughs in China's high-end scientiffc research instruments and equipment.

Keywords: Machine Learning Potential; Chemical Large Language Model; Spectroscopic Large Model;