As business environments are complicated and rapidly changing, preciseness and agility in solving various decision problems faced by a manager are more important. To effectively support the managers problem-solving tasks, a decision support system (DSS) should integrate models for the problems and solvers for the solving algorithms in its system platform, and enable the manager to mix and match the models and solvers freely. Specifically, the manager should be able to know which solvers are applicable to a particular model and to apply an appropriate solver easily. Also, the DSS should be able to integrate at runtime new models and solvers created as the environments change. This paper proposes a model and solver integration framework in a DSS, which facilitates the managers mix-and-match operation and is adaptable to new models and solvers as well. As a conceptual framework, we use the generic model concept for developing the system constructs and procedures. These constructs and procedures are represented as self-contained and well-modularized objects that can be applied to a wide variety of problem domains.