WeDoctor Secures National Algorithm Filing for Three AI Medical Models Ahead of IPO Filing

On November 1, the Cyberspace Administration of China publicly released the eighth batch of algorithm registration information for deep synthesis services. WeDoctor’s independently developed “WeDoctor Medical Large Language Model,” “WeDoctor Medical Assistant Large Language Model,” and “WeDoctor Health Assistant” successfully passed the algorithm registration. Securing algorithm registration for these three large language models signifies that WeDoctor’s LLMs have received official recognition for their technical capabilities in natural language processing and their service capabilities within the healthcare industry.

Figure | WeDoctor’s Three Large-Model Algorithms Pass the Eighth Batch of Filings with the Cyberspace Administration of China

The algorithm filing for deep synthesis services with the Cyberspace Administration of China (CAC) represents a significant regulatory measure in China aimed at ensuring the compliance, security, and transparency of algorithms used in internet information services. It serves as a national-level authoritative endorsement of the security and regulatory compliance of large models concerning data, content, and algorithms. As an internationally leading medical AI platform, WeDoctor has been strategically positioned in the “AI + Healthcare” sector for many years, consistently dedicated to transforming cutting-edge technologies into practical medical service capabilities. The three large models that have recently obtained algorithm filing all demonstrate its technical advantages in delivering precise and personalized healthcare management services.

Industry experts stated that in the highly regulated healthcare sector, where standards are stringent and entry barriers are exceptionally high, securing three filings at once is not only an official endorsement of WeDoctor’s technical prowess but also a reaffirmation of its leadership position in the field of medical AI.

Notably, the WeDoctor Medical Large Language Model has distinguished itself on CMB (Comprehensive Medical Benchmark in Chinese), a leading domestic benchmarking platform for medical large language models, by securing the top position with an impressive score of 91.71. This achievement fully demonstrates its leading capabilities in core functions such as medical knowledge comprehension and clinical decision support. Underpinning these accomplishments is WeDoctor’s continuous deepening of scenario-based applications and technological iterations. By developing diverse business scenarios, including the AI Health Community and AI Hospitals, WeDoctor has not only amassed vast amounts of high-quality data resources but also continuously optimized the model’s adaptability and practicality, enabling it to serve the public’s health needs more precisely and efficiently.

Micro Medical Group’s large healthcare AI model has not only achieved remarkable technological breakthroughs but also demonstrated significant efficacy advantages in practical applications. In the Health Community projects co-developed with local governments such as Tianjin, four intelligent agents—“AI Physician,” “AI Pharmacist,” “AI Health Manager,” and “AI Intelligent Control”—developed on Micro Medical Group’s robust large healthcare AI model foundation have been deployed at scale. This has established a closed-loop application for full-lifecycle health management covering pre-diagnosis, during-diagnosis, and post-diagnosis stages. Supported by AI technology, healthcare services have rapidly improved in quality and efficiency. The Tianjin Health Community now serves over 18,000 patients per day on average, with nearly 200,000 chronic disease patients benefiting from efficient chronic disease management services, leading to a significant reduction in per capita medical expenses.

Furthermore, WeDoctor’s “AI Health Manager” agent has improved the work efficiency of health managers by 150%, while its “AI Pharmacist” has raised the rationality and cost-effectiveness of medication use to over 98%. Following the implementation of the “AI Doctor,” the quality of primary care treatment plans has increased by 22%, and the annual primary-care screening rate for related tests has risen by 28%. Additionally, “AI Intelligent Control” has boosted medical and health insurance compliance by 9.6%.

In terms of data accumulation, the WeDoctor platform has amassed over 400 million dialogue records, 200 million medical case files, and 200,000 physician diagnosis and treatment entries. After standardized de-identification, these datasets provide a solid foundation for the training and effective application of AI models. By implementing various strategies such as alignment algorithms, supervised fine-tuning for role-based dialogues, and reinforcement learning from human feedback (RLHF), WeDoctor’s large medical language model demonstrates outstanding performance in multimodal tasks, including natural language processing and image recognition, thereby significantly enhancing user interaction experience and service quality.

As the national medical consortiums transition and upgrade into health consortiums, the health consortium model in regions such as Tianjin is becoming a “benchmark paradigm” for the new direction of healthcare reform. A representative from WeDoctor stated that the company will continue to deepen the nationwide promotion of its AI-enabled health consortium model, actively respond to the policy guidance of national healthcare reform, advance the strategic upgrade to Artificial Intelligence Hospital 3.0, expand the spectrum of diseases covered by AI-assisted diagnosis and treatment, and explore more innovative chronic disease management models. The recent registration of three algorithms will undoubtedly inject stronger momentum into WeDoctor’s development, facilitating higher-level AI empowerment and delivering richer, more efficient solutions to the healthcare industry.