The Stereotype Content Model and Behaviors from Intergroup Affect and Stereotypes model suggest that perceptions of others’ warmth and competence significantly influence emotional responses and behavioral tendencies. While these models have been successfully applied to human-robot interaction with results consistent with interpersonal interactions, the role of interactive context remains underexplored. This study examines how different interactive contexts—instrumental versus interactional—influence college students’ perceptions of robot warmth and competence. Through an online experiment with 151 participants viewing video prototypes of human-robot interactions, we found that context significantly affects perception outcomes. Instrumental contexts enhanced both warmth and competence perceptions compared to interactional contexts. High-warmth, high-competence robots elicited admiration and facilitation behaviors, while low-warmth, low-competence robots evoked contempt and harm behaviors. However, ambivalent emotions (envy and pity) showed weaker effects than predicted by the original models. These findings underscore the importance of context in HRI research and provide insights for designing context-appropriate robot behaviors.