Under incomplete information, a game model is used to investigate the influence of ownership level and learning ability on the stability of technology innovation alliance from the perspective of knowledge transfer. The decision-making processes of involved parties are divided into two stages in the model. In the first stage, the firm possessing advanced technology decides on the level of knowledge it transfers to its alliance partner. In the second stage, the decision of the parties on whether to maintain or terminate the alliance is based on two factors: the level of knowledge learned and profits gained. The outcomes of the Cournot---Nash equilibrium in the model can reveal when the parties decide to maintain or terminate the alliance. The model explores the status of alliance stability under different ownership levels and learning abilities to provide theoretical support for the selection of optimal dynamic competitive-cooperative relationship and managerial flexibility.