In response to the problems on profit model and transaction pricing of distributed energy storage providing multiple customized power services for sensitive customers, a multi-grade pricing strategy for distributed energy storage to provide various customized power services is proposed, such as reactive power compensation, voltage sag control and harmonic control. Firstly, based on the four-quadrant operation characteristics of energy storage converter, a multi-grade evaluation indicator system of customized power services is established considering the differentiated user demand to power quality, and a cost-to-capacity model for energy storage to provide customized power services in different power quality standards is constructed. Secondly, by taking economic loss of power quality into account, a user customized power utility function is established with individual rational constraint. In consideration of power quality default risk and investment cost constraints, a customized power revenue model of distributed energy storage is constructed. Furthermore, a multi-grade trading framework for distributed energy storage to provide differentiated customized power services and its multi-grade pricing optimization strategy are proposed. Finally, in order to obtain the optimal additional tariff and user purchase package for premium power, the nonlinear multi-grade pricing model is transformed into a mixed integer linear programming model for optimization by using the big M method and transforming the user utility function into a constraint. The comparative analysis of the algorithm verifies that the proposed strategy can reduce the annual cost of customized power services for users while enhancing the energy storage revenue.