Purpose On the basis of congruity theory and language expectancy theory, this study aims to investigate the effect of artificial intelligence (AI) tour guides’ conversation style and destination type congruity on consumers’ continuance usage intention. Design/methodology/approach Three scenario-based experiments were conducted with a total of 818 participants recruited via Credamo, a widely used platform with established reliability in prior research. Findings The results reveal that when AI tour guides adopt a warm conversation style for hedonic destinations or a competent style for utilitarian destinations, consumers’ continuance usage intention for the AI tour guide is significantly enhanced. Trust partially mediates this relationship, while language typicality and trust together form a serial mediation path, shedding light on the underlying mechanism of the congruity effect. Originality/value This study enriches the literature on language design cues in AI tour guides by addressing the research gap concerning the congruity between AI tour guide conversation style and contextual fit.