In the last decades caching has become the key technology used to reduce the performance gap through memory hierarchies across temporal or spatial locales; in particular, the effect is prominent in disk storage systems. Applications that involve heavy I / O activities, which are common in the cloud, are likely to benefit more from caching. The use of local volatile memory as a cache could be a natural alternative, but many well-known constraints, such as the capacity and use of host machines, make it difficult to use effectively. In addition to the technical challenges, the provision of cloud caching services faces an important practical problem (quality of service or agreement level of agreement) of prices. Currently (public) cloud users are limited to a small set of uniform, coarse grain services such as High-Memory and High-CPU on Amazon EC2. In this paper we present the cache model as a service (CaaS) as an optional service to typical infrastructure service offerings. Specifically, the cloud provider reserves a large pool of memory that can be dynamically partitioned and assigned to standard infrastructure services such as disk cache. We first investigate the feasibility of providing CaaS with the concept proof elastic cache system (using dedicated remote memory servers) built and validated in the current system; and the practical benefits of CaaS for users and suppliers (ie, yield and profit, respectively) are studied in depth with a new pricing scheme.