Over the last years, diverse modeling languages have been developed to facilitate multi-level modeling in cell biology. However, using these languages still involves considerable simulation efforts, which hampers their wider applicability to simulation studies. Our aim is to develop more efficient execution strategies that take properties and requirements of multi-level models into account. To reach this goal, methods for on-demand configuration, parallelization, and approximation will receive particular attention. The research will be based on ML-RULES, a rule-based multi-level modeling language realized in the plug-in based modeling and simulation framework JAMES II, which will be refined and extended as part of the project. The impact of the developed methods in supporting multi-level modeling and simulation will be evaluated by both performance experiments and user studies. The results of the project will enrich the portfolio of modeling languages, execution methods, and evaluation strategies in the area of multi-level modeling and simulation.