Wind-induced interference effects occur between high-rise twin buildings. To elucidate the impact of aerodynamic optimization measures on the interference effects between high-rise buildings, large eddy simulations (LES) were conducted to analyze the wind load characteristics of buildings with varying spacing. The simulation results were compared with experimental data and previous studies to validate the method's accuracy. To investigate the influence of various aerodynamic corner optimizations on interference effects between high-rise buildings, LES was employed to simulate wind load characteristics. Simulation results were compared with experimental data and previous studies to ensure the method's accuracy. Three corner treatments-chamfered, rounded, and recessed-were applied to the model, with a corner modification rate of 10%. Under wind field conditions corresponding to a Reynolds number (Re) of 22,000, 10 simulated schemes with spacing ratios (B/L) ranging from 1.2 to 8 were considered, where B represents the center-to-center distance between two square cylinders, and L denotes the side length of each square cylinder. The aerodynamic coefficients, flow field mechanisms, and vorticity of the two side-by-side square cylinders under three corner treatments were compared. Additionally, peak factors were calculated based on predefined measurement points to further assess differences in interference factors. Using the Lorentz function, the peak fitting curve of the wind pressure probability density was compared with the Gaussian fitting curve. The analysis indicates that the wind pressure shielding effect is most pronounced for the rounded corner square cylinder, while the amplification effect is strongest for the chamfered corner square cylinder. At critical spacing, the adjacent facades of the recessed corner square cylinders exhibited a bimodal distribution in wind pressure probability density.