m-WildVision Multilingual Multimodal Large Model Evaluation Dataset

The m-WildVision dataset is a multilingual and multimodal large model evaluation dataset released by Cohere For AI in 2025. It contains 23 languages, each with 500 challenging user query instances from the WildVision-Arena platform. The structure of the dataset includes question ID, language type, instruction text, and image data, and is designed to evaluate the generalization and robustness of the model in different languages.