Yun Zhisheng (09678) released the “Shanhai‧Zhi Medical Grand Model 5.0” version of the medical field expert model

Zhitong Finance App News, Yun Zhisheng (09678) issued an announcement. Recently, the company released the medical expert model “Shanhai Zhi Medical University Model 5.0” version (“Medical Big Model”) to build a “Medical Text Big Model+Medical Multimodal Big Model” dual-core product system, which comprehensively covers core capabilities such as text processing, intelligent collaboration, and multi-modal perception, integrating multiple functions such as medical knowledge questioning, clinical task planning, and image analysis to form one of the most versatile medical AI technology support systems in the industry. Large medical models can deeply simulate clinical thinking for complex reasoning, and achieve accurate differential diagnosis and evidence-based decision-making. The company relies on massive high-quality medical data and exclusive knowledge base training to reach the highest level in the industry in core tasks such as medical knowledge questioning and information extraction; through continuous course learning, knowledge map enhancement, and evidence-based alignment training, medical illusions are minimized and the authenticity and verifiability of the output content is ensured.

In December 2025, in MedBench 4.0, an authoritative domestic Chinese medical model evaluation system built by the Shanghai Institute of Artificial Intelligence, the company achieved excellent results of 94.6 points, 65.7 points, and 59.6 points in medical intelligence, big language models, and multi-modal models, respectively. It took first place in all three technical models of medical models, and used strength to set an industry benchmark. Meanwhile, in the “MedaiBench Evaluation List (Excellent Domestic Medical Model)” published by Zhejiang, the National Artificial Intelligence Application Pilot Center (Healthcare), the company's medical model is honorably listed with solid technology accumulation, demonstrating its hard-core strength in the AI medical field.

The company's medical model has been implemented on a large scale in many top three hospitals, and has cooperated with the top three hospitals to complete technical verification in scenarios with extremely high professional requirements such as clinical-assisted diagnosis and treatment. Its technical capabilities are deeply adapted to medical industry standards and actual clinical needs, fully demonstrating its clinical practical value and industry adaptability.