This paper developed a software named Lexical Complexity Analyzer for Japanese Texts in Multiple Registers (LCA-JT-MR), which is built on MeCab, a widely utilized tokenization and part-of-speech tagging system for Japanese. This software offers a straightforward interface and user-friendly operation, enabling the automatic, efficient, and objective analysis of lexical complexity in both general and academic Japanese texts. It also supports analysis of native Japanese words, Chinese-origin words, and verbs. To illustrate the practical application and effectiveness of this software in Japanese L2 writing research, we employed it to study the lexical complexity of Chinese learners’ Japanese doctoral dissertations on Japanese language research. The results revealed that the differences in lexical density, lexical compositionality, and lexical diversity between learners’ doctoral dissertations and native speakers’ journal articles were relatively small. However, lexical sophistication in the former was significantly higher than in the latter. From the application in this study, it is evident that the employment of the software significantly broadens the scope of Japanese L2 lexical complexity research and substantially enhances the efficiency and objectivity of data processing.