Abstract: Distributed renewable energy is becoming popular in cities, and the use of energy storage is an effective measure to cope with the uncertainty of renewable energy. An optimization model for the operating cost of a shared energy storage system, which is provided by the service of a third-party agent, is proposed based on generalized Nash bargaining. The operating cost of each subject in the Nash equilibrium of the master-slave game is taken as the disagreement point, and a Nash bargaining model of buildings and energy storage agent is established. The model can be transformed into two sub-problems: minimization of the operating cost and pricing bargaining of virtual energy storage. In the sub-problem of virtual energy storage bargaining, the electricity of charging and discharging is used to measure the contribution of buildings and is used as a bargaining power parameter to negotiate with the energy storage agent to set the leasing price of virtual energy storage jointly. The simulation result shows that buildings and energy storage agent have equal bargaining rights under the Nash bargaining, which results in lower leasing price, thus the operating cost of the buildings can be reduced, the revenue of the energy storage agent can be enhanced, and the global optimal economic efficiency of the shared energy storage system can be achieved. Moreover, compared with the general Nash bargaining, the method, which is based on bargaining power, can reflect the difference in the contribution of each subject. Therefore, the enthusiasm of buildings participating in energy storage sharing can be stimulated and the payback period of investment of energy storage agent can be shortened.