In the past few years, the AI healthcare industry has had an awkward problem: the story is big, the scene is real, but commercial realization is always slow.

Many companies can talk about "medical big models," "data intelligence," "real-world research," and "AI hospitals," and can also make good pilot projects. However, once it comes to financial reports, the market is most concerned about three issues: whether revenue can grow? Can gross profit be improved? Can the losses really stop?

The 2026 fiscal year performance released by Yidu Technology provides a relatively clear answer.

As of the fiscal year ending March 31, 2026, Yidu Technology achieved a revenue of 819 million yuan, a year-on-year increase of 14.6%; Gross profit of 298 million yuan, a year-on-year increase of 24.8%; The annual profit was 78.766 million yuan, with a net profit attributable to the parent company of 72.709 million yuan, achieving a key turnaround since its listing and even in the company's development process. More noteworthy is that the company's net cash outflow from operating activities narrowed by 91.2% year-on-year, and the operating cash flow in the second half of the year has turned positive, reaching 29.3 million yuan.

This is not just a "good-looking" financial report, but a report that makes the market re evaluate Meddu Technology.

Because it proves not that a certain business suddenly exploded, but that the company has finally moved from the "investment period" to the "realization period" in the long-term track of AI healthcare.

What truly exceeds expectations is not only profitability, but also the quality of profitability

In April of this year, Yidu Technology released a positive profit forecast, expecting a net profit of 55 million to 70 million yuan for the 2026 fiscal year and a net profit attributable to the parent company of 41 million to 56 million yuan; At that time, the market had already regarded it as an important signal for the company's first full year profit.

But the final result was stronger than the forecast: the annual profit was 78.766 million yuan, exceeding the upper limit of the previous profit forecast by about 12.5%; The net profit attributable to the parent company was 72.709 million yuan, exceeding the previous upper limit of the profit forecast by about 30%. This is the true meaning of 'exceeding expectations' - it's not about going from losses to small profits, but rather the cautious expectations originally given by the management, which were actually exceeded by the final report.

More importantly, this profit was not achieved through cutting R&D, selling assets, or one-time gains. From the financial report, it can be seen that Yidu Technology did not choose to sacrifice long-term technological investment for short-term profits, but instead promoted multiple dimensions such as revenue recovery and growth, gross profit margin improvement, expense ratio reduction, improvement of accounts receivable and impairment pressure, and cash flow repair simultaneously.

In the fiscal year 2026, the overall gross profit margin of the company increased from 33.5% to 36.4%. Among them, the gross profit margin of the AI for Medical sector has significantly increased from 29.7% to 39.7%, and the explanation of the financial report mainly comes from AI efficiency improvement and project refinement management. Sales and marketing expenses as a percentage of revenue decreased from 19.5% to 16.1%, administrative expenses as a percentage of revenue decreased from 20.5% to 8.1%, and research and development expenses as a percentage of revenue decreased from 26.1% to 20.1%.

Behind these numbers, there is actually a signal from a CFO's perspective that is very important: Meddu Technology is not simply pursuing revenue scale, but consciously improving the quality of every piece of revenue.

Especially for an AI healthcare company, the quality of profitability is more important than profitability itself. Because AI healthcare is not a consumer Internet, you can't buy traffic by burning money; Medical institutions, pharmaceutical companies, payers, and government clients all highly value delivery capability, stability, and credibility. Only when the product truly enters the business process, will revenue become sustainable; Only when delivery efficiency improves, can gross profit margin continue to improve.

This is also the most noteworthy aspect of this financial report: the profit of Yidu Technology comes from the business model starting to shift towards "less heavy asset investment, more product reuse, and higher delivery efficiency".

The Long Term Narrative of Yidu Technology: From "Data Processing" to "Medical AI Operating System"

The most difficult thing for the market to understand about Yidu Technology in the past was what kind of company it was.

It is not traditional SaaS, not just CRO, not insurance TPA, nor is it a medical big model startup company. Its business spans across hospitals, pharmaceutical companies, insurance, patients, and government governance, making it easy for outsiders to see it as' doing everything '. But an important signal released by this financial report is that Yidu Technology is consolidating its previously seemingly scattered businesses onto a common foundation.

This base is YiduCore.

According to the financial report, YiduCore is not a single model, but an intelligent base that integrates data governance platform, medical knowledge system, disease model, large model capabilities in medical vertical fields and real scene application framework. As of March 31, 2026, YiduCore has processed and analyzed nearly 9 billion authorized real-world medical records, covering over 1.32 billion patients. The specialized disease database covers over a hundred diseases, and the disease knowledge graph basically covers known diseases. It has also supported the construction of 21 specialized disease domain datasets.

There is a deeper change in this passage: MedTech is redefining itself as a "medical AI operating system" company.

The so-called "operating system" does not mean that Yidu Technology aims to replace hospital systems, pharmaceutical company systems, or insurance systems, but rather that it attempts to deposit the most fundamental and difficult to reuse capabilities of medical AI: data governance, medical knowledge, disease models, clinical pathways, evidence generation, intelligent agent scheduling, and real-world scenario feedback. Then, based on the same set of capabilities, output different products to different customers.

This is also the key to understanding the three business lines of Yidu Technology.

AI for Medical is aimed at hospitals and regulatory agencies, delivering intelligent capabilities at both the hospital and regional levels, including AI middleware, clinical assistant Copilot, medical intelligent circulation, medical intelligent management, high-quality datasets, and more. It addresses efficiency issues in doctor diagnosis and treatment, hospital research, operational management, and regional governance.

AI for Life Sciences provides research and evidence generation capabilities for pharmaceutical companies, device companies, and clinical research clients, including real-world research, patient screening, clinical trial quality control, medical strategy analysis, and post market evidence support. It solves the problems of long R&D cycles, difficult patient recruitment, and high evidence generation costs for pharmaceutical companies.

AI for Care targets insurance, payment providers, and patient populations, providing health management and payment side operational capabilities, including welfare insurance operation, intelligent claims, chronic disease management, digital therapy, and proactive health services. It addresses issues of payment efficiency, risk management, and population health intervention.

On the surface, these are three business lines; Essentially, it is a commercial interface of the same base in different scenarios of "medical, pharmaceutical, insurance, and patient".

This is also the difference between Yidu Technology and single point AI application companies. Single point AI applications typically solve a process, and the founder of Meddu Technology chose a slower, heavier, and more difficult to understand path from day one - focusing on disease accumulation and reusable assets, allowing the same disease understanding ability to create value simultaneously in hospitals, pharmaceutical companies, insurance, and patients.

Once this set of disease capabilities is implemented, it can be reused multiple times: supporting doctor decision-making and research on the hospital side, supporting clinical research and evidence generation on the pharmaceutical company side, supporting risk identification and claims management on the payment side, and supporting long-term health management on the patient side.

This is the core of the long-term value of Yidu Technology: it is not selling an AI tool or just doing a project, but trying to turn disease understanding capabilities into medical AI infrastructure that can be reused across scenarios.

So, what is truly worth paying attention to in Yidu Technology's financial report this time is not just its first profit, but its beginning to make its business logic clearer: the three business lines are not a combination of businesses, but a continuous output of capabilities from the same medical AI base to hospitals, pharmaceutical companies, payers, and patients. Whether this base can be reused continuously is the key to determining the long-term value of the company.

The most significant change in this financial report is the shift from "proving oneself capable" to "proving oneself capable of making money"

AI healthcare has long been in an awkward stage: technically, people believe it will definitely change healthcare; In business, people always doubt when it can make money.

The significance of Yidu Technology's 2026 fiscal year performance report lies in its connection between these two issues.

It proves several things:

Firstly, AI healthcare is not limited to pilot and conceptual levels. Yidu Technology has provided solutions to 133 top Chinese hospitals and 54 regulatory agencies and policy makers, covering over 12000 hospitals; The life sciences sector serves 98 clients, with 17 out of the top 20 multinational pharmaceutical companies being its clients; The AI for Care platform has over 22 million active users.

Secondly, the commercialization of AI healthcare does not rely on single point explosive products, but on workflow penetration. The truly valuable part of Yidu Technology is the AI platform Copilot、 The composite capabilities formed by the integration of Yidu Zhixun, Yidu Zhiguan, high-quality datasets, real-world research, Huiminbao operation, and chronic disease management.

Thirdly, the management began to speak with financial results rather than conceptual narratives. The combination of revenue growth, increased gross profit margin, positive operating profit, higher than expected net profit attributable to shareholders, improved cash flow, buybacks, and dividends creates a new corporate image: it is no longer just a company that talks about the future of AI healthcare, but is beginning to use financial reports to prove that AI healthcare can become a sustainable business.

That's also why this financial report deserves a re examination of Yidu Technology.

The capital market is most afraid of 'AI companies that only know how to tell stories'. But if an AI healthcare company can embed technology into real healthcare scenarios, convert scenarios into revenue, convert revenue into gross profit, convert gross profit into profit and cash flow, and then give back some of the results to shareholders, then it has crossed a critical threshold.

The key word for Yidu Technology's fiscal year 2026 is not just 'first-time profit'. More precisely, AI healthcare is entering the era of flowering and fruiting.