The medical vertical sector has witnessed remarkable strides in artificial intelligence applications, heralded by the official debut of Baichuan-M4, a cutting-edge, medical-augmented large-scale model. This model has exhibited outstanding prowess in the HealthBench medical assessment, clinching the premier position not only on the overall leaderboard but also topping the Hard and Professional sub-leaderboards, thereby demonstrating medical acumen that outshines GPT-5.5. The crux of its advancement lies in revolutionizing interaction patterns, boasting the capability to proactively pose follow-up inquiries and emulate the diagnostic reasoning of genuine physicians, thus elevating the efficiency of consultations and the precision of diagnoses.

Moreover, Baichuan-M4 introduces a 'Full-Course Disease Memory' feature, which seamlessly integrates patients' historical data, markedly enhancing long-context clinical memory performance in comparison to its predecessors. Concurrently, the model integrates 'evidence anchoring' technology, enabling medical conclusions to be meticulously aligned with specific excerpts from authoritative medical literature or clinical guidelines, achieving an evidence-based citation accuracy rate of up to 90.0%, significantly surpassing that of comparable large-scale models.

This launch signifies a monumental leap in the utilization of large-scale models within the medical realm, transitioning from a mere 'repository of knowledge' to a 'diagnostic decision support' system. It furnishes physicians in medically underserved regions with dependable diagnostic references. Furthermore, it presents a novel paradigm for the inclusive deployment of 'AI + healthcare,' poised to propel the intelligent transformation of the medical industry and render precision healthcare more universally attainable.