As conversational agents become increasingly prevalent in educational contexts, it remains unclear how text-based chatbots and embodied social robots differ in terms of user experience. This study addresses that precisely that, using a conversational agent called Bookbot, which is designed to have book discussions with 10- to 11- year-olds. Bookbot was applied to two interfaces, one text-based chatbot interface and the other a social robot interface called Furhat. Ten university students was recruited were they had one interaction with each interface. Employing a mixed-methods design, quantitative data were collected via the User Experience Questionnaire (UEQ) and qualitative insights through thematic analysis through semi-structured interviews. Although no statistically significant differences emerged on any UEQ scale, descriptive trends indicated superior performance of the chatbot on pragmatic dimensions (perspicuity, efficiency, dependability) and overall attractiveness, whereas the robot excelled on novelty and stimulation. Thematic analysis revealed that the chatbot’s clarity and predictable pacing fostered ease of comprehension and control, while the robot’s face and voice generated stronger social presence but suffered from turn-taking breakdowns and speech delivery issues. These findings highlight complementary strengths in the two interfaces: chatbots deliver functional stability and clarity, while social robots enable richer social engagement. Accordingly, design implications include integrating minimalist social cues into chatbots and enhancing the conversational robustness of social robots. Future research should employ issue‐based metrics and longitudinal social‐presence measures to distinguish novelty effects from sustained user engagement.