Jingyi Qianxun Medical AI Model 2.0 Launches with Enhanced Diagnostic Accuracy in Complex Clinical Scenarios

At the JD Cloud City Conference held on May 20, Wang Guoxin, Chief Scientist of JD Health Exploration Research Institute (JDH XLab), announced that “Jingyi Qianxun,” JD Health’s self-developed large medical language model, will receive a major upgrade to version 2.0, following its full open-source release in February this year. This comprehensive upgrade aims to leverage technological innovation and ecosystem openness to drive medical artificial intelligence from generalist services toward deeper expansion into more complex, specialized disease areas, injecting new intelligent momentum into clinical decision-making and patient health management.

Currently, the development of medical artificial intelligence is facing systemic opportunities and challenges. Traditional AI models rely on static data and struggle to simulate the dynamic reasoning capabilities required for clinical diagnosis; the inherent uncertainty in medicine conflicts with AI’s pursuit of deterministic answers, posing higher demands on the technology.

In this regard, JD Health believes that the ultimate goal of large medical models is not to “list all possibilities,” but to perform stepwise reasoning based on evidence-based medicine principles and reasonable information collection, thereby achieving high-probability differential diagnoses and treatment recommendations. For instance, by leveraging multimodal perception (medical history, imaging, and laboratory reports) and deep retrieval agents, large models avoid excessive listing and directly output clinically actionable conclusions, significantly enhancing practical utility. This philosophy provides a new standard for the research, development, and evaluation of medical AI.

Based on this underlying logic, the newly upgraded “Jingyi Qianxun” Medical Large Model 2.0 integrates multi-source clinical information through multimodal perception and deep retrieval technologies, building a decision support system that closely mirrors real-world scenarios. Meanwhile, by leveraging patient data in conjunction with evidence-based medicine databases and clinical case repositories, it strikes a balance between the determinism of AI and the flexibility of medical decision-making, providing physicians with more reliable auxiliary tools and significantly enhancing clinical practicality.

In the realm of specialized disease treatment, “Jingyi Qianxun” 2.0 demonstrates breakthrough value. Addressing complex scenarios in conditions such as malignant tumors and cardiovascular and cerebrovascular diseases, it provides clinicians with personalized diagnosis and treatment recommendations by deeply integrating evidence-based medicine with individual patient data. Its imaging and pathology diagnosis module can rapidly analyze imaging and pathological images, significantly reducing physicians’ workload in image interpretation and aiding treatment decision-making. Meanwhile, JD Health has partnered with leading academic and medical institutions in China to foster industry-academia-research collaboration, driving the translation of technology from the laboratory to the clinical frontline and accelerating the widespread adoption of medical AI.

Since its comprehensive open-source release in February this year, dozens of hospitals, health management centers, pharmaceutical retail enterprises, physical examination centers, and elderly care institutions have begun leveraging the open-source “Jingyi Qianxun” medical large language model to drive service upgrades and operational model transformation. The “Jingyi Qianxun” large language model has made significant progress in achieving scaled, full-scenario applications within the healthcare industry.

Meanwhile, leveraging the “Jingyi Qianxun” medical large language model as its technological foundation, JD Health continues to refine its comprehensive product suite, “AI Jingyi,” designed for all online scenarios. By simulating roles such as physicians, pharmacists, nutritionists, and psychological counselors, JD Health has launched specialized AI agents—including AI Physicians, AI Pharmacists, AI Nutritionists, and AI Psychological Counselors—accelerating their application in areas such as intelligent consultation, medication guidance, online follow-ups, and patient education. To date, over 80% of physician consultation records on JD Internet Hospital have utilized AI assistance. The AI substitution rate for JD Health’s AI Nutritionists has exceeded 90%, with the AI conversion rate among consumers purchasing maternal and infant nutritional products surpassing 45%, outperforming human customer service representatives.

As the industry’s first large-model product designed for comprehensive hospital scenarios, “JD Zhuoyi” provides hospitals with a personal healthcare concierge for patients, digital twins for physicians, and “Future Digital Hospital” services for hospital operations and management. Currently, this product has been pioneered at the First Affiliated Hospital of Wenzhou Medical University. The closed-loop outpatient patient service workflow driven by large models, jointly developed by both parties, has served over 1.8 million patient visits to date.

For users and the general public, JD Health continues to leverage the public-benefit value of its service offerings—such as the intelligent health assistant “Kangkang,” the AI psychological companion “Liaoyu Xiaoyuzhou,” and multimodal technology applications like “mobile blood pressure measurement”—to accelerate the development of an online medical service ecosystem gateway.

The major upgrade to “Jingyi Qianxun” 2.0 marks a significant milestone for JD Health in the field of medical AI. By focusing on specialized disciplines and diseases, strengthening evidence-based reasoning, and expanding application scenarios, JD Health is driving the deep integration of medical AI technologies and advancing industry progress. This initiative provides higher-quality, more efficient healthcare solutions for patients, physicians, and medical institutions, while also offering new perspectives and innovative leadership for the research and development of medical AI and the practical implementation of large language model evaluations.

In the future, JD Health will continue to deepen its collaborations with hospitals, pharmaceutical companies, and research institutions, exploring the potential of AI applications in healthcare. By leveraging AI, we aim to promote the equitable distribution of medical resources, ensuring that high-quality healthcare services benefit a broader population.