The article describes the process of detecting similar texts from a database of multilingual full texts: changing the text format, morphological analysis of words in the text and translating them using a dictionary, comparing two texts based on a set of words. The definition of multilingual similarity of texts can be used in the process of searching for information from full texts related to bibliographic records in information-library systems, automatic replenishment of the electronic library stock based on abstract data from Internet resources, automatic determination of keywords in cataloging, plagiarism of documents in hidden and translated forms in anti-plagiarism systems.