The planning of a multi-family-residential building can be characterized as a multi-objective optimization problem, as multiple criteria must be considered and negotiated with each other. The output of this process - the building design - has a great impact on the social, economic, and ecological performance of the building for a long time. Therefore, it is crucial to find the optimal solution with the best trade-off between the performance criteria. The challenges that the planner faces during the planning process are (a) a large number of the design options and (b) that the fitness criteria (goals) vary from project to project and even during the work at one design. Currently, the process of the manual solution search is highly time-consuming, and hence, it is usually not possible to guarantee that the "true" optimal solution is found. The goal of this project is to develop a method for efficient automatic generation of design options for multi-family-residential buildings. The focus of the project is the cross-scale generation of designs, from the building envelopes to the detailed floorplans.