Mobile devices suffering from resource poverty, using the outside resources can help to cover the resource limitations of mobile devices. Mobile Cloud Computing (MCC) provides data storage and processing services on the cloud rather than on the mobile devices themselves. So, MCC is a possible solution to overcome mobile device resource limitations. Although the offloading to the cloud is a promising solution, sometimes mobile users could not be able to connect to the cloud to get services and faced several obstacles, including; network connectivity, latency challenges, energy consumption, and security. The cloudlet concept was introduced to cover the drawbacks of direct offloading to a remote cloud, but mobile users lose the free mobility when using cloudlet as they must be in the transmission range of it. So, offloading to the cloud and cloudlet is not always the best solution. As the rapidly increasing in mobile devices' memory and its computational capabilities, it could be a computational service provider. Also, the development of wireless peer-to-peer (P2P) communication model encourages to enhance MCC model by offloading tasks to nearby mobile devices. In this paper, we propose M.Cloudlet model, using Wi-Fi Direct to connect nearby mobile devices in an ad-hoc network to establish a mobile cloudlet. M.Cloudlet can provide some cloud services and can be alternative to request services from remote cloud servers or central servers of stationary cloudlet. We implemented and evaluated M.Cloudlet model, and the result shows using M.Cloudlet to offload tasks to nearby devices reduces execution time, therefore, gains more performance and decreases battery consumption.