艾瑞咨询近日发布《中国医疗大模型行业研究报告》，东软集团凭借添翼医疗大模型的落地实践，成为报告中唯一实现双重维度标杆定位的市场主体。报告指出，医疗大模型的竞逐早已脱离参数榜单的竞速，归于产业落地价值的深度兑现。此次东软既被视作头部医疗IT企业向AI原生IT体系演进的范式样本，也获得独立专题篇章深度解构，为观察医疗AI产业化进程提供了重要参照。
报告将国内医疗大模型产业演进划分为四大周期：参数竞赛阶段逐步退潮，数据壁垒构建窗口持续收窄，场景深度嵌入周期加速崛起，产业模式重构的新格局初露端倪。这一划分显示，行业正从单一技术维度比拼，转向对业务场景和全链路效率的深度改造。同时，市场选型逻辑发生根本性迁移，决策主体由技术研发团队转向医疗机构业务运营端，价值评判标尺也从单点准确率指标，迭代为端到端任务完成率、业务全链路效率增益等复合型实绩标准。
在行业评价体系全面重构的背景下，东软添翼医疗大模型何以成为报告唯一双维度聚焦的标杆范例？关键在于其立足既有业务底座，探索大模型与医疗业务的深度融合。面对如何突破“医疗IT系统提供商”固有标签的时代考题，东软正通过落地实践重塑行业认知，稳步推进品牌战略升级。
当前市场主要分化出两种落地思路。一类厂商将AI作为业务系统之外的“外挂能力”，以外部集成方式附着于现有业务系统，虽然落地较快，但与原有业务流程和数据体系相对割裂，容易造成跨系统操作、数据孤岛和业务体验不一致等问题。另一类思路以东软新一代添翼为实践，着力打造AI原生的医院业务底座，不是简单在成熟系统上外挂AI，而是以AI重新定义业务系统的能力架构。
具体而言，东软将大模型、智能体、知识、数据与业务能力原生融合，并深入HIS、EMR等核心业务流程，使AI成为业务系统的内生能力，让医生在原有业务场景中自然获得智能辅助，实现业务与智能的一体化演进。这一路径以“模式共振”为核心，强调AI原生的本质不只是把AI嵌入业务，而是让数据、知识、模型与业务场景形成持续反馈和持续进化的闭环。
在技术架构上，东软添翼AI3.0内生AI一体化平台以“添添智能中枢”为核心，搭建“医疗大模型+专科模型+智能体生态”的多层次架构；依托“数据基座—赋能平台—价值扩展”体系，完成多源异构医学数据治理与知识萃取，把高质量医疗数据转化为模型迭代燃料，让大模型在临床场景中持续演进。
这一案例折射出医疗大模型赛道竞争已进入新阶段：模型性能不再是差异化来源，能否真正嵌入核心业务流程、创造可量化的效率增益，将决定企业在下一轮产业格局重构中的身位。随着场景深度嵌入周期加速到来，医疗大模型的价值兑现路径正变得更加清晰。
iResearch recently released the China Medical Large Model Industry Research Report, and Neusoft Corporation, backed by the real-world deployment of its Tianyi medical large model, has become the only market player to secure a dual-dimension benchmark positioning in the report. The report notes that competition in the medical large model arena has long moved beyond parameter leaderboards and is now centered on delivering deep industrial value. Neusoft is recognized both as a paradigm of a leading healthcare IT company evolving into an AI-native IT architecture, and as the subject of a dedicated deep-dive chapter, offering an important reference for understanding the industrialization of medical AI.
The report divides the evolution of China's medical large model industry into four major phases: the parameter competition stage is gradually receding, the window for building data moats continues to narrow, the cycle of deep scenario embedding is accelerating, and a new landscape of industrial model reconstruction is beginning to emerge. This classification indicates that the industry is shifting from competition on a single technological dimension to deep transformation of business scenarios and end-to-end efficiency. Meanwhile, the market's technology selection logic has fundamentally changed: decision-making has shifted from technical R&D teams to business operations teams at healthcare institutions, and the yardstick for evaluation has evolved from single-point accuracy to composite performance metrics such as end-to-end task completion rates and full-chain business efficiency gains.
Against the backdrop of a sweeping reconfiguration of industry evaluation criteria, why has Neusoft's Tianyi medical large model become the sole benchmark case to receive dual-dimension focus in the report? The key lies in its ability to build on an existing business foundation and explore the deep integration of large models with healthcare operations. Facing the era-defining challenge of transcending its long-held label as a "healthcare IT system provider," Neusoft is reshaping industry perception through real-world implementation and steadily advancing its brand strategy upgrade.
The current market has broadly diverged into two implementation approaches. One group of vendors treats AI as a "bolt-on capability" outside the business system, attaching it to existing systems through external integration. Although this allows faster deployment, it remains relatively disconnected from the original business processes and data architecture, often leading to cross-system operational friction, data silos, and inconsistent business experiences. The other approach, exemplified by Neusoft's new-generation Tianyi, focuses on building an AI-native hospital business foundation. Rather than simply bolting AI onto mature systems, it uses AI to redefine the capability architecture of business systems.
Specifically, Neusoft natively fuses large models, intelligent agents, knowledge, data, and business capabilities, and embeds them deep into core business processes such as HIS and EMR. This makes AI an intrinsic capability of the business system, allowing physicians to receive intelligent assistance naturally within their existing workflows and enabling a converged evolution of business and intelligence. This path is built around "mode resonance," emphasizing that the essence of AI-native is not merely embedding AI into business, but creating a closed loop in which data, knowledge, models, and business scenarios continuously feed back and evolve together.
In terms of technical architecture, the Neusoft Tianyi AI 3.0 Native AI Integration Platform, centered on the Tiantian Intelligent Hub, establishes a multi-tier architecture of "medical large model + specialized models + intelligent agent ecosystem." Relying on a "data foundation – enablement platform – value expansion" framework, it performs governance and knowledge extraction across multi-source heterogeneous medical data, converting high-quality medical data into fuel for model iteration and enabling the large model to continuously evolve in clinical scenarios.
This case reflects that competition in the medical large model track has entered a new stage: model performance is no longer a source of differentiation. Whether a solution can truly embed itself in core business processes and deliver quantifiable efficiency gains will determine a company's position in the next round of industrial restructuring. As the cycle of deep scenario embedding accelerates, the path to value realization for medical large models is becoming increasingly clear.