In recent years, a variety of approaches have been developed for the structure prediction of protein-peptide complexes, and they have been increasingly incorporated into peptide drug design [1]. Very recently, we saw a breakthrough in protein structure prediction: the AlphaFold AI-based method demonstrated high accuracy competitive with experimental techniques. First attempts have been made to use the AlphaFold methodology in the prediction of protein-peptide complexes. In my talk, I will discuss the current status of the structure prediction field in application to protein-peptide complexes and future coming prospects. In addition, I will provide examples of our modeling pipelines allowing for protein-peptide docking with significant flexibility of peptides as well as protein receptors [2, 3].