The advancement of artificial intelligence (AI) has led to the development of various tools applied in multiple fields, from process automation to decision making. However, the performance evaluation of these tools remains a key challenge to ensure their effectiveness, accuracy and efficiency. In this context, this paper performs a literature review based on the SALSA framework, evaluating conversational AI tools under certain criteria, also offering a general comparison of their operation and performance in different scenarios and areas, considering metrics such as accuracy, processing speed, resource consumption and generation capacity, highlighting that, although the results show the performance of each tool, it varies according to the context of use. The results obtained seek to contribute to the development of more robust and reliable AI systems, promoting their optimal implementation in productive and scientific environments.