Optimizing urban spatial form has become an important research topic for promoting urban sustainable development and improving energy efficiency. This study selects 164 high-rise residential blocks in the Changsha area as the research object and constructs three multi-objective optimization frameworks and mathematical models for the spatial form of high-rise residential blocks based on the Rhino and Grasshopper parametric platform. The TOPSIS comprehensive evaluation method is combined to further screen the Pareto front solutions, and global sensitivity analysis of spatial form parameters is conducted using Simlab to explore the multi-objective optimization strategies for the spatial form of high-rise residential blocks in the Changsha area, guided by energy performance. Practical cases are selected for verification. The results show that, (1) after multi-objective optimization, the optimal scheme for the determinant type reduces EUI by 7.22%, increases PVP by 28.84%, and reduces CGR by 2.21%. The optimal scheme for the point-cluster type reduces EUI by 3.80%, increases PVP by 7.34%, and reduces CGR by 2.30%. The optimal scheme for the staggered type reduces EUI by 6.66%, increases PVP by 17.65%, and reduces CGR by 2.26%. (2) The potential order of spatial form optimization for the three types of high-rise residential blocks in the Changsha area is determinant type > staggered type > point-cluster type. (3) The spatial form parameters most sensitive to the three architectural performance goals of EUI, PVP, and CGR are floor area ratio and average number of floors. The research results will provide quantifiable references for scientific decision-making in the design practice of high-rise residential blocks in the Changsha area and even in regions with hot summers and cold winters.