@inproceedings{zhang-etal-2026-realchart2code,
    title = "{R}eal{C}hart2{C}ode: Bridging the Gap in Real-World Chart-to-Code Generation via Multi-Task Evaluation",
    author = "Zhang, Jiajun  and
      Li, Yuying  and
      Li, Zhixun  and
      Guo, Xingyu  and
      Wu, Jingzhuo  and
      Zheng, Leqi  and
      Yang, Yiran  and
      Zhang, Jianke  and
      Li, Qingbin  and
      Yan, Shannan  and
      Jia, Changguo  and
      Wu, Junfei  and
      Wang, Zilei  and
      Liu, Qiang  and
      Wang, Liang",
    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/check-for-anonymous-pdfs/2026.acl-long.1945/",
    pages = "41995--42032",
    ISBN = "979-8-89176-390-6",
    abstract = "Vision-Language Models (VLMs) have demonstrated impressive capabilities in code generation across various domains. However, their ability to replicate complex, multi-panel visualizations from real-world data remains largely unassessed. To address this gap, we introduce \textbf{RealChart2Code}, a new large-scale benchmark with over 2,800 instances grounded in authentic datasets and featuring tasks with clear analytical intent. Crucially, it is the first benchmark to systematically evaluate chart generation from large-scale raw data and assess iterative code refinement in a multi-turn conversational setting. Our comprehensive evaluation of 14 leading VLMs on RealChart2Code reveals significant performance degradation compared to simpler benchmarks, highlighting their struggles with complex plot structures and authentic data. Our analysis uncovers a substantial performance gap between proprietary and open-weight models and confirms that even state-of-the-art VLMs often fail to accurately replicate intricate, multi-panel charts. These findings provide valuable insights into the current limitations of VLMs and guide future research directions."
}