A dataset of 1,460 authentic Russian sentences extracted from the Russian National Corpus and independently annotated with CEFR levels by 12 certified RFL experts, along with a collection of prompts for LLM-based automatic sentence complexity assessment.

RU-CEFR-Short is dataset of short CEFR-graded fragments of Russian nonfiction texts (15-30 words), extracted from Russian L2 textbooks.

Collection of word lists with frequencies

A Parallel Russian-Simple Russian Dataset

Word alignments from Russian-Simple Russian parallel data