To realize lightweight design of complex tensegrities with multi self-stress modes, a double-stage mass optimization in conjunction with quantum beetle search algorithm is proposed. In the first stage, based on the form-finding theory of tensegrity structure, the L2-norm is employed to find the integral feasible prestress. In the second stage, the mathematical optimization model, by setting minimum structural mass under full stress as the objective function, is established with the considerations of constraint conditions regarding cables and bars. The adjustment of the prestress level and the Similarity Transformation Strategy (STS) are originally promoted to guarantee the continuity and efficiency of the proposed algorithm. Three illustrative examples, including a four-way tensegrity grid, double-layer open reticulated tensegrity shells and single layer tensegrity plates, are comprehensively studied to demonstrate the effectiveness of the proposed method in lightweight design of tensegrities with multi self-stress modes.