FinMMEval
FinMMEval introduces the first multilingual and multimodal evaluation framework for financial large language models, assessing models’ abilities in understanding, reasoning, and decision-making across languages and modalities to promote robust, transparent, and globally inclusive financial AI systems.
Tasks
The FinMMEval Lab features three different tasks:
Financial Exam Question Answering
Evaluates models’ conceptual understanding and domain reasoning using multilingual, professional exam-style financial questions (e.g., CFA, CPA, EFPA, BBF). Performance is measured by accuracy.
Multilingual Financial Question Answering
Tests analytical financial reasoning using multilingual and multimodal information sources (e.g., SEC filings + cross-lingual news). Models generate concise, evidence-grounded answers evaluated with ROUGE and factuality metrics.
Financial Decision Making
Assesses reasoning-to-action capabilities by generating Buy/Hold/Sell trading decisions and short rationales based on textual and numerical market contexts (BTC, TSLA). Evaluated via profitability, stability, and risk metrics (e.g., Sharpe Ratio, Cumulative Return).
Organizers
- Zhuohan Xie (Mohamed bin Zayed University of Artificial Intelligence, UAE)
- Rania Elbadry (Mohamed bin Zayed University of Artificial Intelligence, UAE)
- Fan Zhang (The University of Tokyo, Japan)
- Georgi Georgiev (Sofia University “St. Kliment Ohridski”, Bulgaria)
- Xueqing Peng (The Fin AI, USA)
- Lingfei Qian (The Fin AI, USA)
- Jimin Huang (The Fin AI, USA)
- Dimitar Dimitrov (Sofia University “St. Kliment Ohridski”, Bulgaria)
- Vanshikaa Jani (University of Arizona, USA)
- Yuyang Dai (INSAIT, Bulgaria)
- Jiahui Geng (Mohamed bin Zayed University of Artificial Intelligence, UAE)
- Yuxia Wang (INSAIT, Bulgaria)
- Ivan Koychev (Sofia University “St. Kliment Ohridski”, Bulgaria)
- Veselin Stoyanov (Mohamed bin Zayed University of Artificial Intelligence, UAE)
- Preslav Nakov (Mohamed bin Zayed University of Artificial Intelligence, UAE) "