FinMMEval

The inaugural edition of FinMMEval, a CLEF 2026 Workshop dedicated to the multilingual and multimodal evaluation of financial AI systems. Real-world financial decision-making relies on diverse data sources, including textual reports, visual documents, and time-series signals, spread across languages and modalities. To support robust, interpretable, and auditable AI in this domain, FinMMEval proposes three complementary pilot tasks: (1) Document Parsing and Structured Extraction from scanned financial filings, (2) Financial Exam Question Answering, and (3) Financial Decision Making based on market context, combining historical prices, news, and portfolio status. The lab includes datasets in six languages, such as English, Spanish, Arabic, Greek, Bulgarian and Hindi, covering low-level perception to high-level decision-making. We aim to foster cross-disciplinary collaboration across NLP, CV, time-series modeling, and financial industries, and establish FinMMEval as a benchmark for the next generation of financial AI systems.

Organizers

• Zhuohan Xie (Mohamed bin Zayed University of Artificial Intelligence)
• Rania Elbadry (Mohamed bin Zayed University of Artificial Intelligence)
• Georgi Georgiev (Sofia University “St. Kliment Ohrdiski”)
• Xueqing Peng (Yale University)
• Jimin Huang (The FinAI)
• Yuxia Wang (Mohamed bin Zayed University of Artificial Intelligence)
• Dimitar Dimitrov (Sofia University “St. Kliment Ohrdiski”)
• Vanshikaa Jani (Mohamed bin Zayed University of Artificial Intelligence)
• Ivan Koychev (Sofia University “St. Kliment Ohrdiski”)
• Veselin Stoyanov (Mohamed bin Zayed University of Artificial Intelligence)
• Preslav Nakov (Mohamed bin Zayed University of Artificial Intelligence)