In recent years, anthropomorphism has attracted extensive attention from academia and industry as a critical factor in chatbot service design. As an important extension of anthropomorphism, zoomorphism—integrating animal images, behaviors, or traits into design—has not yet been thoroughly explored regarding its influence mechanism in service failure scenarios. Based on the Stereotype Content Model and Mind Perception Theory, this study examines the impact of two types of zoomorphic chatbots (warmth-oriented vs. competence-oriented) on consumers' service failure tolerance, and reveals the mediating role of experience and agency as well as the moderating mechanism of demand uncertainty. The results show that:(1)Compared with competence-oriented zoomorphic chatbots, warmth-oriented ones can significantly enhance consumers' service failure tolerance;(2)Experience plays a mediating role between zoomorphic chatbot types and service failure tolerance, while the mediating effect of agency is not verified;(3)Demand uncertainty significantly moderates the impact of zoomorphic chatbot types on both experience and agency. This study expands the theoretical understanding of human-machine interaction and service failure response from the perspective of zoomorphism, and provides practical implications for enterprises to design and select appropriate chatbot types and mitigate the negative impacts caused by service failures.