In recent years, the acceptance and use of intelligent robots and other kinds of intelligent systems have begun to gain more and more attention also in information systems research. Here, many studies have found the perceived intelligence of robots to act as one critical antecedent for their acceptance and use, but few studies have focused on the antecedents of perceived intelligence itself. In this study, we aimed to address this gap in prior research by examining the effects of individual intelligence dimensions on the overall intelligence perception of robots in the work context. In addition, we also examined the potential differences in these effects as well as in the individual intelligence dimensions and overall intelligence perception themselves between three common types of robots: physical robots, software robots, and chatbots. These examinations were based on online survey data from 1,080 present or prior users of robots at work. In summary, we found that adaptability, personality, autonomy, and multifunctionality act as the most influential antecedents of perceived intelligence in the case of all three types of robots. In addition, we also found that software robots and chatbots perform better than physical robots in most individual intelligence dimensions and in overall intelligence perception.