As the popularity of cloud computing rises, both the number of decision criteria relevant for the cloud consumers and the number of cloud service offers increase. The high number of options makes it challenging to determine which cloud services to purchase while satisfying all requirements and maintaining a low cost. In industry, these decisions are often taken on an ad-hoc basis, without systematic procedures, making it difficult to optimize the decisions and reducing costs. Previous research introduced multiple approaches to make automated and cost-optimal decisions on which cloud service provider to select and which cloud services to purchase. However, each approach only solves parts of the problem, such as quality of service requirements or uncertainties of usage predictions. If some requirements cannot be modeled and the approach is not extendable, it is not usable in practice as the proposed solution might violate some requirements. Thus, a tool is needed that can combine multiple decision criteria and pricing models, while being extendable with future requirements. This thesis aims to close this gap by proposing a framework for cloud cost optimization. Decision criteria for the optimization problem, such as quality of service requirements and pricing models, can be formalized via mixed integer programming and combined into one problem description. A prototype has been implemented in Python and evaluated for plausibility, expressiveness, and practical usability.