MedXpertQA is a high-quality multimodal dataset in the medical field, released by Tsinghua University and Shanghai Artificial Intelligence Laboratory in 2025. The relevant paper results are:MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding", which aims to provide an authoritative evaluation benchmark for medical question answering, medical image understanding and clinical reasoning models.

The dataset contains 4,460 sample data, integrating text and image data, covering tasks such as medical question answering, clinical diagnosis, treatment plan recommendation, and basic medical knowledge understanding. It supports the research and development of complex medical decision-making capabilities and is suitable for fine-tuning and evaluation of medium-scale models in the medical field.