Soft robots are made of compliant materials that perform their tasks by deriving motion from elastic deformations. They are used in various applications, e.g., for handling fragile objects, navigating sensitive/complex environments, etc., and are typically actuated by Pneumatic/hydraulic loads. Though demands for soft robots are continuously increasing in various engineering sectors, due to the lack of systematic approaches, they are primarily designed manually. This paper presents a systematic density-based topology optimization approach to designing soft robots while considering the design-dependent behavior of the actuating loads. We use the Darcy law with the conceptualized drainage term to model the design-dependent nature of the applied pressure loads. The standard finite element is employed to evaluate the consistent nodal loads from the obtained pressure field. The robust topology optimization formulation is used with the multi-criteria objective. The success of the presented approach is demonstrated by designing a member/soft robot of the pneumatic networks (PneuNets). The optimized member is combined in several series to get different PneuNets. Their CAD models are generated, and they are studied with high-pressure loads in a commercial software. Depending upon the number of members in the PneuNets, different output motions are noted. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.