Tairex said it introduced a system on Wednesday, Oct. 21, that brings AI agents from different medical specialties into one review of complex cases so physicians can weigh their findings together.
The Multidisciplinary Human-Agent Collaboration Ecosystem is designed for cases that cross specialties. Tairex says it can activate agents at the same time and show their evidence and disagreements in a shared interface. Doctors are expected to assess the recommendations and retain responsibility for critical medical decisions, the company said. Tairex says it was incubated by the Institute for AI Industry Research at Tsinghua University.
The company illustrated the process with a patient who had palpitations and a history of asthma. An endocrinology agent considered a beta-blocker, while an agent for respiratory medicine flagged possible bronchoconstriction from a nonselective drug and suggested considering a selective beta-1 option. The example came from Tairex’s announcement.
Tairex says Agent Hospital includes 54 AI doctors across more than 20 specialties and tens of thousands of diseases. The company says training involved simulations with millions of AI patients, and that one agent can conduct tens of thousands of simulated consultations per day. Tairex reported accuracy above 96% on MedQA.
The announcement gives no date for the system’s introduction in practice and names no hospitals that will use it. Tairex says its broader Agent Hospital platform began a public beta at the end of November 2025 with eight hospitals in Beijing and Guangxi; its beta notice says human doctors make medical decisions. The company has presented primary-care settings as a possible beneficiary of broader specialist access, but did not identify a clinic using this system.
