Abstract Transformable models are 3D models whose shapes can be changed by rotating or translating their component parts. They have a variety of applications in our daily lives, being used in film props and sets, robots, furniture, tools, and toys. Successful transformable models, however, are challenging to create. In this paper, we present a new approach to designing and animating a transformable model, in which a source model is optimally segmented based on a target model and skeleton provided by users, and the motion of transformation is mapped from the source to target models. Our experimental results indicate that our system can transform a 3D model plausibly.