Purpose <p>Large language models have demonstrated significant potential across a wide range of medical fields and may potentially induce a transformation in the field of infertility and sterility. This systematic review aims to synthesize existing research and explore the current strengths and limitations of large language models in the field of infertility and sterility.</p> Methods <p>Researchers conducted a comprehensive search across three databases and employed a thematic synthesis approach for data analysis.</p> Results <p>The analysis included a total of 13 studies. Large language models exhibited advantages in the accuracy and reproducibility of information output and consultation, demonstrated robust learning capabilities, and were able to provide satisfactory recommendations to patients. However, there were significant variations in the performance of different large language models, and the readability of the output information was poor, making it difficult to provide comprehensive answers.</p> Conclusion <p>In the domain of infertility, large language models have not yet achieved full reliability and should be regarded solely as information sources whose outputs require careful verification; they are presently incapable of substituting for clinical diagnosis. Their development requires increased investment in relevant technologies and the use of authoritative, accurate information to unlock their potential in this field and other medical areas.</p>