At today's media briefing, Wang Xiaochuan, founder and CEO of Baichuan Intelligence, officially launched the new open-source medical-enhanced large language model Baichuan-M3. The model achieved the top comprehensive performance in multiple authoritative medical professional evaluations, not only surpassing OpenAI's latest release GPT-5.2, but also outperforming the average level of human practicing physicians in some tasks, marking a key breakthrough of domestic large models in high-barrier vertical fields.

Baichuan-M3 is deeply optimized for medical scenarios, integrating a vast amount of medical literature, clinical guidelines, real medical records, and drug knowledge bases. It has precise disease reasoning, medication suggestions, test interpretation, and patient communication capabilities. Its open-source strategy aims to promote the co-construction of the medical AI ecosystem, accelerating the application of technology in grassroots medical care, auxiliary diagnosis, and health management scenarios.

Wang Xiaochuan reflected on his entrepreneurial journey, stating that when Baichuan Intelligence was founded in 2023, many people questioned whether "AI could truly enter the medical field." However, with the emergence of specialized models like DeepSeek in 2025, the acceptance of AI by the medical community significantly increased; and the fact that international giants such as OpenAI and Anthropic have successively launched HIPAA-compliant medical assistants further proves that "AI in healthcare has moved from proof of concept to large-scale application," becoming one of the core directions for current AI implementation.

He also specifically mentioned Ant Group's recent launch of the medical health AI "Afu," believing that this reflects how healthcare, as the "crown jewel" of AI, is rapidly moving from professional laboratories to everyday life. Whether it is chronic disease management, medication reminders, or initial symptom screening and health consultation, AI is seamlessly integrating into the health management loop of hundreds of millions of users in a safe and accessible way.

Wang Xiaochuan emphasized that Baichuan Intelligence will continue to focus on the medical vertical field, promoting AI to become doctors' "super assistants" and patients' "health gatekeepers" through open source and industry collaboration. In 2026, when AI and life sciences are deeply integrated, the release of Baichuan-M3 is not only a technological leap, but also an important milestone for Chinese AI power to participate in global competition in medical intelligence.