<p>This study develops a stochastic tripartite evolutionary game model involving governments, banks, and high-emission enterprises to examine green credit policy effectiveness. I calibrate the model using Chinese data and analyse equilibrium dynamics under two scenarios: (i) weak regulation with green credit and innovation, and (ii) strong regulation with green credit and innovation. Results reveal three key findings. First, while both subsidies and penalties influence system evolution, subsidies prove more effective in promoting bank participation and enterprise innovation. Second, potential conflicts emerge between environmental regulation and green credit mechanisms in promoting green technology adoption; however, coordinated regulatory enforcement can mitigate these tensions and guide enterprises toward innovation. Third, in complex institutional environments characterized by high corruption, government-enterprise interactions exhibit cyclical cooperation-defection patterns, though well-timed policy interventions can break these cycles and facilitate green transitions. This study offers insight into how to classify government actions to create external motivations for green enterprise innovation.</p>