Core Viewpoint - The AI medical company Yidu Tech (02158) has demonstrated business resilience and improved profitability in its interim results for the fiscal year 2026, with significant revenue growth and a substantial reduction in losses [1][2] Financial Performance - Total revenue reached RMB 358 million, representing an 8.7% increase compared to the same period last year [1] - Losses decreased sharply by 72%, nearing breakeven [1] Business Segments - The AI for Medical segment generated revenue of RMB 153 million, a year-on-year increase of 14.6%, with a customer base expanding to 127 well-known hospitals and 44 regulatory bodies [1] - The AI for Life Science segment reported revenue of RMB 138 million, maintaining high customer loyalty with a coverage rate of 17 out of the top 20 multinational pharmaceutical companies [1] - The AI for Care segment saw a revenue surge of 30.3% to RMB 66.67 million, with the "Shenzhen Huimin Insurance" program reaching over 6.15 million insured individuals [1] Strategic Focus - The company’s strategic focus and refined operations have led to steady revenue growth, significant loss reduction, and improved cash flow [2] - The ample cash reserves and comprehensive ecosystem layout support ongoing innovation and the pursuit of sustainable profitability in the evolving AI medical industry [2]

Group 1 - The company Yidu Technology (医渡科技) reported interim results for the six months ending September 30, 2025, with revenue of approximately 358 million yuan, an increase of 8.7% year-on-year; gross profit was about 134 million yuan, up 14.2% year-on-year; and the loss narrowed to 15.76 million yuan, a decrease of 72% year-on-year [1] - The big data platform and solutions segment generated revenue of 153 million yuan, a year-on-year increase of 14.6%, primarily benefiting from expanded business opportunities driven by policy support [1] - The health management platform and solutions segment reported revenue of 66.6 million yuan, a year-on-year increase of 30.3%, mainly due to changes in the product mix [1] Group 2 - YiduCore, the company's core algorithm engine, continued to advance the research and training of large language models in the medical vertical during the reporting period, training on over 500 billion high-quality tokens and analyzing nearly 7 billion authorized medical records from over 1.3 billion patient visits, creating a dataset system covering more than 98 specific diseases [2] - The model won the "Best Paper Award" and the championship in the "Medical NLP Code Auto-Generation Evaluation" at the 11th China Health Information Processing Conference (CHIP 2025) and also secured first place in the medical health track at the 2025 "Data Element ×" competition in Jiangxi [2] - As an important ecological partner in the national artificial intelligence application pilot base in Beijing's medical field, YiduCore achieved large-scale deployment across medical, pharmaceutical, and health insurance sectors through a "one base, multiple applications" model, fully implementing a comprehensive solution across "medicine - pharmaceuticals - insurance - patients" [2] Group 3 - During the reporting period, the company provided solutions to 127 top hospitals in China and 44 regulatory agencies and policymakers, covering over 10,000 hospitals in total [3] - The company utilized a dual-driven approach of "data elements × AI technology" to solidify its data intelligence foundation and accelerate the large-scale application of intelligent solutions [3] - The large model technology assisted multiple hospitals in upgrading their data governance and research platforms, with clinical research expanding into oncology, respiratory, and traditional Chinese medicine disciplines, while the doctor Copilot product matrix achieved full coverage of the diagnosis and treatment process, with over 50,000 uses in a single hospital [3]

Group 1 - The company reported total revenue of RMB 358.1 million for the six months ending September 30, 2025, representing a year-on-year growth of 8.7% [1] - Revenue from the big data platform and solutions increased by 14.6%, while revenue from the health management platform and solutions grew by 30.3%, driven by policy support and expanded business opportunities [1] - The company's loss narrowed to RMB 15.8 million, a decrease of 72.0% year-on-year, indicating improved operational efficiency and cost control [1] Group 2 - The core algorithm engine, YiduCore, advanced the development of large language models in the medical vertical, utilizing over 500 billion high-quality tokens and analyzing nearly 7 billion authorized medical records [2] - YiduCore won the championship in the "Medical NLP Code Automatic Generation Evaluation" and the "Best Paper Award" at the 11th China Health Information Processing Conference (CHIP 2025) [2] - The company provided solutions to 127 top hospitals in China and 44 regulatory agencies, covering over 10,000 hospitals, and supported significant upgrades in data governance and research platforms [3] Group 3 - The health management platform maintained a leading position in core cities, serving as the main operating platform for "Shenzhen Huimin Insurance" for three consecutive years and "Beijing Huimin Insurance" for five years [3] - The number of insured individuals under "Shenzhen Huimin Insurance" exceeded 6.15 million, with a cumulative service of over 18 million people in three years, reflecting increasing brand loyalty [3] - Active users on the health management platform who completed at least one transaction surpassed 22 million [3]

Core Insights - The company reported total revenue of RMB 358.1 million for the six months ending September 30, 2025, representing an 8.7% year-on-year increase, driven by significant growth in its big data and health management platforms [1] - The company narrowed its losses to RMB 15.8 million, a 72.0% reduction compared to the previous year, while cash flow from operating activities saw a 56.0% year-on-year decrease in net outflow [1] - YiduCore, the company's core algorithm engine, has made significant advancements in developing large language models for the healthcare sector, processing over 13 billion patient records [2] Financial Performance - Total revenue reached RMB 358.1 million, with big data platform revenue increasing by 14.6% and health management platform revenue growing by 30.3% [1] - The company’s loss for the period was reduced to RMB 15.8 million, indicating improved operational efficiency [1] - Cash reserves totaled RMB 2.0993 billion, reflecting a strong liquidity position [1] Technological Advancements - YiduCore has trained its large language model using over 500 billion high-quality tokens and has built a dataset covering more than 98 specific diseases [2] - The company won multiple awards for its medical NLP capabilities, including first place in the healthcare track at the "Data Elements ×" competition [2] Market Engagement - The company provided solutions to 127 top hospitals and 44 regulatory bodies, covering over 10,000 hospitals in total [3] - The number of life sciences clients reached 88, with 17 out of the top 20 multinational pharmaceutical companies as clients [3] - The health management platform has maintained a leading position in core cities, with over 6.15 million insured under the "Shenzhen Hui Min Bao" program [3]

Revenue Performance - For the six months ended September 30, 2025, the total revenue reached RMB 358.1 million, representing an 8.7% year-over-year increase[6]. - Revenue from the big data platform and solutions segment was RMB 153.2 million, up 14.6% compared to RMB 133.7 million in the same period last year[6]. - The health management platform and solutions segment saw a significant growth of 30.3%, with revenue increasing to RMB 66.7 million from RMB 51.2 million[6]. - Revenue for the reporting period reached RMB 138.3 million, a decrease of 4.4% year-on-year[20]. - Revenue increased by 8.7% from RMB 329.4 million for the six months ended September 30, 2024, to RMB 358.1 million for the six months ended September 30, 2025, driven by growth in the big data platform and health management solutions segments[32]. - Revenue from the big data platform and solutions segment rose by 14.6% from RMB 133.7 million to RMB 153.2 million, benefiting from expanded business opportunities due to policy support[32]. - Revenue from the health management platform and solutions segment surged by 30.3% from RMB 51.1 million to RMB 66.6 million, primarily due to changes in the product mix[32]. - Customer contract revenue for the six months ended September 30, 2025, was RMB 358,105 thousand, an increase of 8.3% compared to RMB 329,448 thousand for the same period in 2024[68]. Profitability and Loss - Gross profit for the period was RMB 134.0 million, reflecting a 14.2% increase from RMB 117.4 million in the previous year[6]. - Gross profit increased from RMB 117.4 million to RMB 134.0 million, with gross margin improving from 35.6% to 37.4%[34]. - Operating loss narrowed to RMB 15.7 million, a 71.0% improvement from a loss of RMB 54.3 million in the same period last year[6]. - Period loss reduced by 72.0% from RMB 56.4 million to RMB 15.8 million[41]. - Net loss for the period was RMB 15,760 thousand, a substantial reduction from RMB 56,362 thousand in the previous year[68]. - The company reported a basic and diluted loss per share of RMB 0.01 for the six months ended September 30, 2025, compared to RMB 0.04 in the same period of 2024[69]. Cash Flow and Financial Management - The net cash outflow from operating activities decreased by 56.0% year-over-year, indicating improved cash management[8]. - Cash and cash equivalents decreased from RMB 3,309.1 million to RMB 2,099.3 million as of September 30, 2025[42]. - Other income for the six months was RMB 80,916 thousand, slightly down from RMB 82,937 thousand in 2024[68]. - Total operating expenses, including cost of sales, sales and marketing expenses, administrative expenses, and R&D expenses, amounted to RMB 442,731,000 for the six months ended September 30, 2025, slightly up from RMB 435,455,000 in 2024[84]. User Engagement and Market Reach - The active user base on the health management platform exceeded 22 million, with over 615,000 participants in the "Shenzhen Hui Min Bao" insurance program[10]. - The "Shenzhen Huiminbao" program reached over 6.15 million participants, a historical high, with cumulative claims exceeding 390,000[24]. - Active users on the health management platform exceeded 22 million as of September 30, 2025[27]. Research and Development - The company has focused on breakthroughs in AI model development for the medical vertical, investing in a medical large language model with 70 billion parameters, trained on over 500 billion high-quality tokens[12]. - The AI-driven patient recruitment agent based on medical record understanding has significantly improved the efficiency of clinical trial recommendations and data utilization[14]. - The company has developed a lightweight application model that closely matches the performance of larger parameter models, reducing hardware computational requirements[12]. - The company has established a solid partnership network with key medical institutions and research organizations, enhancing its "AI + Medical Insurance" capabilities[21]. Awards and Recognition - The company has been recognized with industry awards, including first prize in the medical health track at the "Data Element X" competition and the best paper award at the CHIP 2025 conference[13]. Corporate Governance and Future Plans - The company has complied with the corporate governance code as per the Hong Kong Stock Exchange listing rules during the reporting period[55]. - The board of directors does not recommend the distribution of an interim dividend for the six months ended September 30, 2025[67]. - The company has no major investments or capital asset plans for the future as of September 30, 2025[46].

健康管理平台和解决方案（HMPS）新增河北和广州两地"惠民保"业务。 机构：光大证券 研究员：付天姿/王贇 事件：近期公司在AI 医疗创新领域取得多项进展，包括深度参与北京市医疗领域国家人工智能应用中 试基地、中标III 期临床研究项目、新增河北和广州两地"惠民保"业务。 深度参与中试基地建设，以核心技术共筑AI 医疗创新高地。9 月12 日，公司作为北京市医疗领域国家 人工智能应用中试基地的核心生态伙伴，深度参与平台建设与场景创新，推动多项标杆项目落地：开发 区域级受试者招募AI智能体，全面提升受试者招募效率与精准度，解决临床研究受试者招募难、招募 慢的核心痛点，已在6 个项目中落地验证；开发北京市AI 应用插件程序，以统一入口方式整合集成多 AI 应用，大幅减轻医疗机构用户操作和系统集成负担。2025 年8 月，国务院发布《关于深入实施"人工 智能+"行动的意见》，文件指出，要优化应用发展环境，布局建设一批国家人工智能应用中试基地。 而后，上海、北京、广州、浙江相继宣布启动医疗领域国家人工智能应用中试基地，我们认为公司基于 医疗大模型技术将深度受益。 生命科学解决方案（LSS）中标III 期临床研究项目， ...

Core Viewpoint - The report by Everbright Securities highlights significant advancements made by Yidu Tech (医渡科技) in the field of AI healthcare, establishing a preliminary "data-algorithm-scenario" closed-loop model that accelerates business implementation across multiple scenarios [1] Group 1: AI Healthcare Innovations - Yidu Tech is deeply involved in the construction of the national AI application pilot base in Beijing, serving as a core ecological partner and driving the implementation of several benchmark projects [1] - The company has developed an "AI recruitment agent" for clinical trial participants, which has been validated in six projects, significantly improving the efficiency and accuracy of participant recruitment [1] - Yidu Tech has also created an AI application plugin for Beijing, integrating various AI tools to reduce the operational and system integration burden on medical institutions [1] Group 2: Clinical Research Projects - In September, Yidu Tech announced winning a bid for the Phase III clinical research project of recombinant human nerve growth factor (SMR001) eye drops, with a total amount of approximately 55.82 million RMB [2] - The company leverages its medical large model to efficiently produce high-quality clinical evidence and disease knowledge graphs, addressing the pain points of long duration, high costs, and low success rates in clinical trials [2] - With a continuously expanding high-quality client base, Yidu Tech is expected to further enhance its market share in the medium to long term [2] Group 3: Health Management Platform - Yidu Tech has recently launched "Hui Min Bao" insurance services in Hebei and Guangzhou, contributing intelligent services to these initiatives [3] - The company has participated in "Hui Min Bao" projects across 5 provinces and 13 cities, serving over 40 million insured users [3] - Utilizing intelligent models and risk prediction capabilities, Yidu Tech provides comprehensive solutions from product design to operational services for urban inclusive insurance, with expectations for continued penetration into high-tier cities [3] Group 4: Investment Outlook - Everbright Securities maintains a "buy" rating for Yidu Tech, expressing optimism about the company's ongoing innovations and ability to implement AI healthcare solutions [3]

智通财经APP获悉，光大证券发布研报称，考虑到公司在AI医疗创新领域取得多项进展，维持医渡科技 (02158)"买入"评级，近期公司在AI医疗创新领域取得多项进展，包括深度参与北京市医疗领域国家人 工智能应用中试基地、中标III期临床研究项目、新增河北和广州两地"惠民保"业务。 风险提示 深度参与中试基地建设，以核心技术共筑AI医疗创新高地 医疗数字化推进不及预期、医疗数字化行业竞争加剧、生命科学解决方案及健康管理平台和解决方案分 部商业化推进不及预期。 9月12日，公司作为北京市医疗领域国家人工智能应用中试基地的核心生态伙伴，深度参与平台建设与 场景创新，推动多项标杆项目落地：开发区域级受试者招募AI智能体，全面提升受试者招募效率与精 准度，解决临床研究受试者招募难、招募慢的核心痛点，已在6个项目中落地验证;开发北京市AI应用插 件程序，以统一入口方式整合集成多AI应用，大幅减轻医疗机构用户操作和系统集成负担。2025年8 月，国务院发布《关于深入实施"人工智能+"行动的意见》，文件指出，要优化应用发展环境，布局建 设一批国家人工智能应用中试基地。而后，上海、北京、广州、浙江相继宣布启动医疗领域国家人工智 ...

Investment Rating - The report maintains a "Buy" rating for Yidu Technology (2158.HK) [4] Core Insights - Yidu Technology has made significant advancements in AI medical innovation, including deep involvement in the National AI Application Pilot Base in Beijing and winning a Phase III clinical research project [1][2] - The company has expanded its health management platform and solutions by launching "Hui Min Bao" insurance services in Hebei and Guangzhou, enhancing its service reach [3] - Revenue and profit forecasts have been adjusted downward due to slower-than-expected progress in the AI digital therapy business, but the long-term market share is expected to improve [4] Summary by Sections AI Medical Innovation - Yidu Technology is a core partner in the National AI Application Pilot Base in Beijing, focusing on building a "data-algorithm-scenario" closed-loop for AI medical innovation [1] - The company has developed AI recruitment agents to improve clinical trial participant recruitment efficiency, successfully implemented in six projects [1] Clinical Research Projects - The company won a Phase III clinical research project for SMR001 eye drops, valued at approximately RMB 55.82 million, which is expected to enhance its market share in the long term [2] Health Management Solutions - The launch of "Hui Min Bao" in Hebei and Guangzhou has attracted over 200,000 participants on the first day, with the company serving over 40 million users across multiple provinces [3] Financial Forecasts - Revenue forecasts for FY2026 and FY2027 have been reduced by 19.4% and 24.9% to RMB 797 million and RMB 896 million, respectively, while FY2028 revenue is projected at RMB 1.007 billion [4][5] - The net profit forecast for FY2026 and FY2027 has been adjusted to -RMB 85 million and -RMB 36 million, with a positive net profit of RMB 8 million expected in FY2028 [4][5]

Core Viewpoint - The article discusses the challenges and potential of AI in healthcare, emphasizing that while AI shows promise in controlled environments, its integration into real clinical settings faces significant hurdles [1][2][4]. Group 1: AI Performance and Clinical Integration - Microsoft's AI diagnostic system scored four times higher than human doctors in a complex case test, highlighting AI's potential in medical diagnostics [1]. - A study from Harvard indicates that simply providing doctors with AI tools like ChatGPT does not improve diagnostic outcomes; optimal results occur when AI analyzes cases first, followed by human input [4]. - The disconnect between AI's technical capabilities and clinical needs is a core issue, as medicine requires both scientific and artistic approaches, which AI struggles to replicate [5]. Group 2: Challenges in Implementation - The integration of AI into clinical workflows is hindered by established practices among doctors, who may find AI tools burdensome rather than helpful [6]. - Medical data infrastructure is crucial for AI's effectiveness; for instance, Yidu Tech invested over $100 million over four years to build a data foundation necessary for AI applications [6]. - Patient trust remains paramount, as evidenced by a survey where none of the 3,000 patients chose hospitals based on AI tools, indicating that patients prefer human doctors over algorithms [6]. Group 3: Democratization of Healthcare - AI's ultimate goal is to democratize access to quality healthcare, as illustrated by a new insurance model in Beijing and Shenzhen that offers affordable premiums and high coverage [7]. - The use of AI in non-critical care settings is being explored to enhance service delivery, particularly in underserved areas [7]. - AI can reduce administrative burdens on doctors, allowing them to spend more time with patients, thus improving overall healthcare delivery [7]. Group 4: Future Collaboration Between Doctors and AI - There is a consensus that AI will not replace doctors; however, those who do not learn to utilize AI effectively may be outpaced by their peers [8]. - Medical education is evolving to include AI collaboration skills, ensuring future doctors can use AI tools while understanding their limitations [9]. - Continuous monitoring and optimization of AI tools are necessary to ensure they are user-friendly and effective for busy healthcare professionals [9].