This paper presents a generic framework for building quo- tient graphs for Stochastic Well-formed Net models by ex- ploiting the symmetries implicitly defined in the model. Two instantiations are presented, one based on static symmetries and the other one based on dynamic symmetries. The sec- ond method can usually deal with partially symmetric sys- tems in a more effective way than the first one. However, in some cases it may result in a larger graph. We present here a new approach that overcomes the weakness of these methods. All these techniques can be used for efficient per- formance analysis of systems.