Existing Web service optimal combination approach is mainly focused on single tasks using “selfish” behavior to pursue optimal solutions. This causes conflicts because many concurrent tasks compete for limited optimal resources, reducing service quality in services. With the best reply function of quantified task conflicts and game theory as bases, this paper establishes a mathematical model depicting the competitive relationship between multitasks and Web service under QoS constraints, and guarantees that every task can obtain optimal utility services considering other task combination strategies. Moreover, an iterative algorithm which reaches the Nash equilibrium is also proposed, and all tasks attain utility optimization under conflicting environments. Experimental analyses show that the approach can considerably enhance the actual utility of all tasks compared with existing Web services combinatorial methods.