In the restaurant industry, the use of artificial intelligent robots is increasing, but robots are still primarily involved in serving tasks. Therefore, there is a need to study customers" perceptions of robots in cooking. In this study, an extended SERVQUAL model is used to analyze the differences in customers" perceptions of service quality and their impact on customer attitudes by the agent of task (robot chef vs. human chef) and task difficulty (low vs. high). The results showed that customers perceived higher assurance for robot chefs than human chefs, and the interaction effects of the agent of task and task difficulty were significant for service quality. Also, service quality perceptions have significant impacts on customer attitude. This study extends the SERVQUAL framework to the context of food service by categorizing the agent of task and task difficulty and provides primary data on food service robots in restaurants that can increase consumer satisfaction and suggest practical measures for related practitioners.