The article examines the current development status and challenges faced in the field of large-scale models. Applications of large models are becoming increasingly mature, extending from the model end to the application end to address industry pain points. However, startups face shortages of technical talent and data resources, as well as competitive pressures from larger companies. Commercialization paths include cloud services for developers and customized solutions for specific industries. The cooling of investment has also increased the difficulty of entrepreneurship.