The language features of the second language (L2) English academic texts written by Indonesian graduate students enrolled in Hungarian higher education are employed in the present study. The study focuses on the level of abstraction and informational density in student assignments in particular. Seven high-stakes essays were collected from seven Indonesian graduate students registered in the faculty of Social Sciences at three different Hungarian universities. Coh-Metrix, a corpus-based computational tool, was used in this study to examine indices of content words and abstraction. As for the comparison between higher and lower proficiency level students, parametric statistics were used to conduct a quantitative analysis of the selected linguistic features, including nouns, verbs, adjectives, adverbs, pronouns, and concreteness and abstraction. The results show that C1 English proficiency level students outnumber B2 students in terms of informational density. Their texts are more abstract than those of B2 students.