Driven by the national strategies of "digital economy" and "new infrastructure", the internet data center (IDC) has developed rapidly in recent years and gradually become an important new load and demand response (DR) resource in the power system. To support the coordinated development of the IDC and power system, this paper proposes a bi-level collaborative planning framework for the IDC and distribution network considering the potential of flexibilities. First, an IDC energy consumption model is established according its data and power load characteristics, and its flexible adjustment potential is analyzed from the perspectives of time migration and reduction of data load. Through in-depth analysis of the information interaction between the IDC and the distribution network, and considering the influence of uncertain factors, this paper proposes a method in which the distribution network operator (DNO) is the upper leader in formulating real-time node electricity prices. The data center operator (DCO) is the lower-level follower to participate in DR and decide optimal site selection and configuration. This method realizes the bi-level collaborative planning of maximizing the profit of the upper level and minimizing the cost of the lower level by fully considering the constraints of the two planning operations and prices. Finally, the rationality and effectiveness of the method proposed are proved by example simulation and comparative analysis. This work is supported by the National Natural Science Foundation of China (No. 52177082).