Instruction-following capability has become a major ability to be evaluated for Large Language Models (LLMs). However, existing datasets, such as IFEval, are either predominantly monolingual and centered on English or simply machine translated to other languages, limiting their applicability in multilingual contexts. In this paper, we present an carefully-curated extension of IFEval to a localized multilingual version named Marco-Bench-MIF, covering 30 languages with varying levels of localization. Our benchmark addresses linguistic constraints (e.g., modifying capitalization requirements for Chinese) and cultural references (e.g., substituting region-specific company names in prompts) via a hybrid pipeline combining translation with verification. Through comprehensive evaluation of 20+ LLMs on our Marco-Bench-MIF, we found that: (1) 25-35% accuracy gap between high/low-resource languages, (2) model scales largely impact performance by 45-60% yet persists script-specific challenges, and (3) machine-translated data underestimates accuracy by7-22% versus localized data. Our analysis identifies challenges in multilingual instruction following, including keyword consistency preservation and compositional constraint adherence across languages. Our Marco-Bench-MIF is available at https://github.com/AIDC-AI/Marco-Bench-MIF.
@article{arxiv.2507.11882,
title = {Marco-Bench-MIF: On Multilingual Instruction-Following Capability of Large Language Models},
author = {Bo Zeng and Chenyang Lyu and Sinuo Liu and Mingyan Zeng and Minghao Wu and Xuanfan Ni and Tianqi Shi and Yu Zhao and Yefeng Liu and Chenyu Zhu and Ruizhe Li and Jiahui Geng and Qing Li and Yu Tong and Longyue Wang and Weihua Luo and Kaifu Zhang},
journal= {arXiv preprint arXiv:2507.11882},
year = {2025}
}