In today’s digital field, smart city construction has become the core strategic direction of global urban development. With the advancement of my country’s new smart cities, the construction of new infrastructure, and the continuous improvement of spatiotemporal big data platforms, urban top-level design also urgently needs comprehensive digitalization requirements from research and analysis to planning and design, and then to management and operation. This paper studies the development history and research status of computational urban design at this stage and analyzes the modules and paths of existing designs. The article believes that the computational urban design model should be based on parametric thinking and a generative design process with algorithms as the core. On this basis, a computational urban design model with the framework of “target input-morphological generation-performance evaluation-analysis and decision-making” is established, aiming to assist decision makers in building a fast-decision-making path and enhance public participation in urban design.