Abstract: This study constructs a mixed-integer programming model for large logistics companies addressing pickup and delivery problems within complex urban road networks. The model comprehensively considers multiple transshipment points, multiple vehicle types, multiple trips, divisible demands, and time windows, aiming to minimize startup and travel costs. To validate the model’s effectiveness, we used the pr01 dataset as our study case. By establishing different test scenarios, we delved into the impact of various factors on model performance. The results indicate that the model considering all five factors outperforms the one considering only four, with a significant reduction in costs. When vehicle sensitivity to dispatch cost is low, strategy involving indivisible demands offer economic and convenience advantages. The multiple transshipment points and multiple trips factors play a critical role in cost control; neglecting them can lead to a substantial cost increase. However, from a load factor perspective, the single trip plan demonstrates optimal performance, contrasting sharply with conclusions drawn from a cost standpoint. This comprehensive model provides robust support for the operational planning and decision-making of large logistics enterprises, helping to reduce operational costs, enhance transportation efficiency, and improve customer satisfaction.