Abstract: The steel industry is a crucial pillar of national economies. China, in particular, is a major global steel producer. However, it faces significant challenges with regard to technological innovation, product competitiveness, and green and low-carbon development. Intelligent manufacturing has emerged as the key factor driving corporate transformation and upgrading in the steel industry. This paper first presents an in-depth analysis of the concept, essence, and architecture of intelligent manufacturing in steel enterprises. A comprehensive framework encompassing four modules—technological foundation, production operations, organizational change, and corporate benefits—is proposed. This framework considers the process-oriented nature, high energy consumption, and high emissions of steel enterprises. It also considers the technological and organizational innovations required to ensure balanced and sustainable development. Steel coil cutting, which serves as a link between raw material processing and customer order fulfillment, is an essential aspect in achieving full-process optimization in steel enterprises. Optimization of steel coil-cutting plans reflects intelligent decision-making and economic goals within an intelligent manufacturing framework. However, many steel enterprises lack efficient optimization solutions for steel coil cutting. This limits processing efficiency and economic benefits, hindering the improvement of enterprise information systems and intelligent manufacturing levels. Therefore, achieving intelligent steel coil cutting, reducing raw material losses during cutting, and realizing real-time monitoring and optimized management of the production process are crucial in transitioning enterprises toward intelligent manufacturing. This paper presents a mathematical model for the steel coil-cutting problem on the basis of the actual cutting processes in factories. This model considers factors such as two-stage constraints, coil material, order requirements, coil-width limitations, and cutting losses. A steel coil-cutting optimization algorithm is proposed with the objective of minimizing cutting losses. This algorithm facilitates the specification of the width range of the cut coils and simultaneously considers the nesting of the sheet and strip orders. It generates initial cutting plans by using a combination of value and random selections. To avoid local optima, the algorithm dynamically adjusts the order values. It iterates through the "value correction–nesting plan generation–demand update" steps until the order demands are met and then outputs the final cutting plan. The effectiveness of the algorithm is verified through random simulation experiments and using experimental datasets from the literature. The results of the random experiments show that the optimized coil utilization rate is over 97%, significantly outperforming traditional manual methods. Moreover, the algorithm considerably reduces the nesting time. On the datasets used, the proposed algorithm also outperforms two other algorithms in terms of cutting utilization. The two-stage cutting method yields lower operational and time costs compared to the four-stage cutting employed by the extant algorithms. With its effectiveness thus validated, the coil-cutting optimization algorithm can be deployed in steel enterprises to better satisfy personalized and diverse market demands during processing. The algorithm affords positive technological, operational, and beneficial effects. It provides a practical path and innovative model for steel enterprises to achieve intelligent manufacturing transformation. Furthermore, it will potentially drive technological progress and industrial upgrades in the steel industry.