Achieving a high quality and cost-effective tests is a major concern for software buyers and sellers. Using tools and integrating techniques to carry out low cost testing are challenging topics for the testing community. In this work we contribute to alleviate the burden by proposing an architecture support for mutati on and coverage criteria based testing. This architecture integrates metaheuristics to derivate effective test sets and uses automated tools to speed up testing activities. In this paper, we describe different components of the testing architecture. We focus on the developed tools performing mutation testing for JAVA code at method-level and the coveragebased testing for C code at function-level. The integration of metaheuristics in these tools is illustrated by using ant colony and tabu search to derive optimal and pruned test sets. By using our testing architecture two cases study to carry out tests respectively for JAVA methods and C-functions are presented and discusued.