Baichuan releases “M2Plus”, an Evidence-Augmented Medical Model Billed as a “ChatGPT for Doctors”
Want to read in a language you're more familiar with?
Baichuan launches M2 Plus, an evidence-augmented medical AI billed as a “ChatGPT for doctors,” targeting safer clinical reasoning and code generation.
On Oct. 22, 2025, Chinese AI lab Baichuan unveiled Baichuan-M2Plus, a healthcare-focused large model and the latest step in its medical stack, alongside an upgraded companion app and open API access. The company positions M2Plus as a “doctor’s ChatGPT,” optimized for clinical reasoning and evidence-grounded answers.
Baichuan says M2Plus introduces a six-source Evidence Augmented Reasoning (EAR) framework that synthesizes original studies, evidence reviews, clinical guidelines, practice know-how, public health education materials, and regulatory real-world data. In internal evaluations, the model reportedly shows a significantly lower hallucination rate than general-purpose LLMs—about three times lower than DeepSeek’s medical product—and “even surpasses” the popular U.S. medical app OpenEvidence on the company’s tests.
The release follows Baichuan’s open-sourcing of Baichuan-M2 in August and is framed as a push toward safer clinical decision support and patient-facing Q&A. With the API now open, hospitals, health platforms and developers can integrate M2Plus into triage assistants, guideline retrieval, coding/claims workflows, and medication counseling, according to the announcement.
