Along with the rapid progress of heterogeneous networks, multiple user terminals can cooperate via local or short networks as a integrate group to communicate via diverse wide area networks in parallel. This brings forth Concurrent Multipath Transport challenge where the multi flow rate allocation is the key issue. A utility-based model to form a strategy of rate allocation is proposed, in which the user’s QoS demands can be described with user utility function. Moreover, a distributed multi flow rate allocation algorithm that converges to the Pareto optimal is developed. The simulation result demonstrates that the proposed algorithm is efficient to allocate rate for cooperative terminals group in heterogeneous networks.