Public offering oversubscribed 1,073 times as medical imaging AI company debuts on HKEX

医疗影像AI公司港交所首日上市，公开招股获1073倍超额认购

HANGZHOU, China

中国杭州

,

，

March 31, 2026

2026年3月31日

/PRNewswire/ -- Hangzhou Diagens Biotechnology Co., Ltd. (HKEX: 02526) today began trading on the Main Board of The Stock Exchange of Hong Kong. Shares closed at HK$202.60, up 104.65% from the offer price of HK$99.00, giving the company a market capitalization of approximately HK$18 billion (US$2.3 billion) at the end of its first trading day..

/PRNewswire/ -- 杭州迪安生物技术有限公司（港交所：02526）今天开始在香港交易所主板交易。股票收盘价为202.60港元，较发行价99.00港元上涨104.65%，公司在首个交易日结束时市值约为180亿港元（23亿美元）。

The IPO raised approximately US$101 million through the sale of 7,999,200 H shares. The public offer tranche was 1,073 times oversubscribed. Huatai Financial Holdings (Hong Kong) acted as sole sponsor.

通过发售7,999,200股H股，此次首次公开募股（IPO）筹集了大约1.01亿美元。公开发售部分超额认购达1,073倍。华泰金融控股（香港）担任独家保荐人。

Diagens is the first company whose core business is built on a medical imaging AI foundation model to list on a major international exchange. It is also the third large-model AI company to go public on HKEX in 2026, following Zhipu AI (2513.HK) and MiniMax (0100.HK).

迪英加是首家核心业务建立在医疗影像AI基础模型的公司，于主要国际交易所上市。这也是继智谱AI（2513.HK）和MiniMax（0100.HK）之后，第三家于2026年在港交所上市的大型模型AI公司。

Solving a Clinical Bottleneck

解决临床瓶颈问题

Chromosome karyotype analysis is an essential step in prenatal diagnosis and IVF treatment, but it has long been constrained by a shortage of trained specialists. In China, qualifying a cytogeneticist can take up to two years, and demand is outpacing supply as assisted reproduction becomes more prevalent..

染色体核型分析是产前诊断和试管婴儿治疗过程中的重要环节，但长期以来，这一领域一直面临专业人员短缺的困境。在中国，培养一名细胞遗传学专家需要长达两年的时间，而随着辅助生殖技术的普及，相关需求正迅速超过供给。

Diagens' flagship product, AI AutoVision

迪亚琴斯的旗舰产品，AI AutoVision

®

®

, cuts turnaround time from the traditional 14 to 30 days down to 4 to 7 days — a tenfold improvement in turnaround efficiency. The software has been designated a Class III Innovative Medical Device by China's National Medical Products Administration (NMPA) and is currently in final registration review..

，将传统的14到30天的周转时间缩短到4到7天——周转效率提高了十倍。该软件已被中国国家药品监督管理局（NMPA）指定为III类创新医疗器械，目前正处于最终注册审查阶段。

'AI is not just cold code — it is the warmth that protects life,' said Dr. Song Ning, Founder and Executive Director of Diagens. 'Diagens will continue to use the certainty of technology to combat the unknowns of disease, and to create genuine social and technological value for all.'

“人工智能不仅是冰冷的代码，更是守护生命的温度。”Diagens创始人兼执行总监宋宁博士表示，“Diagens将继续用技术的确定性对抗疾病的未知，为所有人创造真正的社会价值与科技价值。”

According to Frost & Sullivan, Diagens held a 30.6% share of China's chromosome karyotype analysis market by revenue in 2024, the highest of any single provider, ending years of dominance by imported systems from Zeiss and Leica.

根据弗若斯特沙利文的报告，2024年迪安占中国染色体核型分析市场收入的30.6%，是所有单一供应商中最高的，结束了多年来蔡司和徕卡进口系统主导市场的局面。

A Foundation Model Built for Clinical Use

专为临床应用构建的基础模型

Underpinning the company's full product portfolio is iMedImage

支撑公司全线产品组合的是iMedImage。

®

®

, certified by Frost & Sullivan as the largest medical imaging foundation model in the world by parameter scale, and the first cross-modal medical imaging foundation model to reach commercial deployment globally.

，经弗若斯特沙利文认证为全球参数规模最大的医疗影像基础模型，也是全球首个达到商业部署的跨模态医疗影像基础模型。

The model supports 19 imaging modalities and covers more than 90% of clinical medical imaging scenarios. It powers the company's iMed MaaS (Model-as-a-Service) platform, which allows hospitals to deploy a specialized AI model locally using as few as 200 imaging samples — compressing development timelines from an industry average of three to five years to under 60 days, at over 90% lower cost.

该模型支持19种成像模式，覆盖了超过90%的临床医学成像场景。它为公司的iMed MaaS（模型即服务）平台提供了动力，该平台允许医院使用最少200个成像样本本地部署专用AI模型——将开发时间从行业平均的三到五年压缩到60天以内，成本降低超过90%。

Tencent has already integrated iMed MaaS into its intelligent pre-annotation capabilities..

腾讯已经将iMed MaaS融入其智能预标注能力中。

Strong Financial Momentum

强劲的财务动能

For the nine months ended September 30, 2025, Diagens reported:

截至2025年9月30日的九个月，Diagens报告称：

Revenue of RMB 112 million, up approximately 470% year-on-year

人民币1.12亿元，同比增长约470%

Overall gross margin of 75.9%, up from 42.9% in the prior-year period

总体毛利率为75.9%，高于上年同期的42.9%

Technology licensing gross margin of 96.5%

技术许可毛利率为96.5%

Technology licensing revenue surpassing hardware sales to become the largest single revenue category for the first time

技术授权收入首次超过硬件销售，成为最大的单一收入类别。

The company's business model spans four streams: medical imaging software, medical devices, reagents and consumables, and technology licensing.

公司的业务模式涵盖四个领域：医疗影像软件、医疗器械、试剂和耗材以及技术授权。

Growth Outlook

增长前景

Diagens currently serves more than 400 medical institutions across all 31 provinces of China. With over 12,000 secondary-level and above hospitals nationwide, market penetration remains at an early stage.

迪安目前服务于中国全部31个省的400多家医疗机构。全国范围内二级及以上医院超过12,000家，市场渗透率仍处于早期阶段。

In November 2025, five Chinese government ministries jointly set a target for AI-assisted medical imaging diagnosis to be deployed universally in secondary-level and above hospitals by 2030 — a policy directive that aligns directly with Diagens' commercial expansion plan.

2025年11月，中国五个政府部委联合制定了一个目标，即到2030年将人工智能辅助医学影像诊断普遍部署在二级及以上医院——这一政策指令与Diagens的商业扩展计划直接吻合。

Beyond karyotype analysis, iMedImage

除了核型分析之外，iMedImage

®

®

's coverage of 19 imaging modalities provides a pipeline of addressable clinical applications. The company has obtained CE certification in the European Union and entered an OEM partnership with Nasdaq-listed Bionano Genomics for international distribution.

“对19种成像模式的覆盖提供了可解决的临床应用管道。该公司已获得欧盟的CE认证，并与纳斯达克上市的Bionano Genomics建立了OEM合作伙伴关系，进行国际分销。”

IPO proceeds will fund further development and commercialization of AI AutoVision

IPO所得将用于进一步开发和商业化AI AutoVision。

®

®

, continued investment in the iMedImage

，继续对iMedImage的投资

®

®

platform, and domestic and international market expansion.

平台，以及国内外市场扩张。

SOURCE Hangzhou Diagens Biotechnology Co., Ltd.

来源：杭州迪安生物技术有限公司

21

21

%

%

more press release views with

更多新闻稿浏览量与

Request a Demo

请求演示