@inproceedings{peng-etal-2026-multifinben,
    title = "{M}ulti{F}in{B}en: Benchmarking Large Language Models for Multilingual and Multimodal Financial Application",
    author = "Peng, Xueqing  and
      Qian, Lingfei  and
      Wang, Yan  and
      Xiang, Ruoyu  and
      He, Yueru  and
      Ren, Yang  and
      Jiang, Mingyang  and
      Zhang, Vincent Jim  and
      Guo, Yuqing  and
      Zhao, Jeff  and
      He, Huan  and
      Han, Yi  and
      Feng, Yun  and
      Jiang, Yuechen  and
      Cao, Yupeng  and
      Li, Haohang  and
      Yu, Yangyang  and
      Wang, Xiaoyu  and
      Gao, Penglei  and
      Lin, Shengyuan  and
      Wang, Keyi  and
      Yang, Shanshan  and
      Zhao, Yilun  and
      Liu, Zhiwei  and
      Lu, Peng  and
      Huang, Jerry  and
      Wang, Suyuchen  and
      Papadopoulos, Triantafillos  and
      Giannouris, Polydoros  and
      Soufleri, Efstathia  and
      Chen, Nuo  and
      Deng, Zhiyang  and
      Fu, Heming  and
      Zhao, Yijia  and
      Lin, Mingquan  and
      Qiu, Meikang  and
      Smith, Kaleb E  and
      Cohan, Arman  and
      Liu, Xiao-Yang  and
      Huang, Jimin  and
      Xiong, Guojun  and
      Lopez-Lira, Alejandro  and
      Chen, Xi  and
      Tsujii, Junichi  and
      Nie, Jian-Yun  and
      Ananiadou, Sophia  and
      Xie, Qianqian",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.770/",
    pages = "16934--16963",
    ISBN = "979-8-89176-390-6",
    abstract = "Real-world financial analysis involves information across multiple languages and modalities, from reports and news to scanned filings and meeting recordings. Yet most existing evaluations of LLMs in finance remain text-only, monolingual, and largely saturated by current models. To bridge these gaps, we present MultiFinBen, the first expert-annotated multilingual (five languages) and multimodal (text, vision, audio) benchmark for evaluating LLMs in realistic financial contexts. MultiFinBen introduces two new task families: multilingual financial reasoning, which tests cross-lingual evidence integration from filings and news, and financial OCR, which extracts structured text from scanned documents containing tables and charts. Rather than aggregating all available datasets, we apply a structured, difficulty-aware selection based on advanced model performance, ensuring balanced challenge and removing redundant tasks. Evaluating 21 leading LLMs shows that even frontier multimodal models like GPT-4o achieve only 46.01{\%} overall, stronger on vision and audio but dropping sharply in multilingual settings. These findings expose persistent limitations in multilingual, multimodal, and expert-level financial reasoning. All datasets, evaluation scripts, and leaderboards are publicly released."
}