KINGDEE INT'L (00268): Subscription revenue grew by 20.4% in the first half of 2026, as AI-native products accelerate commercialization.

In the past decade, Kingdee (00268) addressed the issue of "how software transitions from local deployment to the cloud"; now, the company's challenge has become "what will ERP look like in the AI era." By the first half of 2026, Kingdee's answers began to crystallize: traditional SaaS continues to grow, with subscription revenue maintaining over 20% growth, while AI-native products are beginning to enter customer procurement decisions and are growing at a much faster rate than traditional business. From the numbers, two significant developments are occurring simultaneously at Kingdee: on one hand, the subscription business left over from the cloud transformation of the past decade continues to grow and begins to release operational leverage; on the other hand, AI-native business is growing at a rate far exceeding that of traditional SaaS and is starting to make a real contribution to revenue. This is key to understanding Kingdee's mid-2026 performance. Rather than the company searching for a new growth curve, it is more accurate to say that the original SaaS curve is becoming steeper, with AI emerging as a new variable driving this upward trajectory. From Cloud Realization to AI Realization As of the first half of 2026, Kingdee delivered a commendable performance: total revenue reached 3.625 billion yuan, a year-on-year increase of 13.6%, which is faster than the 11.2% growth in the same period last year. Among this, cloud service revenue was 3.116 billion yuan, up 16.6% year-on-year, accounting for 86% of total revenue; subscription revenue stood at 2.028 billion yuan, a year-on-year increase of 20.4%, with a share exceeding half at 55.9%; and the subscription ARR reached 4.413 billion yuan, an 18.3% year-on-year growth. Notably, revenue from AI-native products reached 296 million yuan, representing a staggering year-on-year growth of 189%, with AI-native subscription revenue reaching 146 million yuan, up 282.9%; the total value of related signed contracts was 605 million yuan, increasing by 159.2%. Meanwhile, the company achieved a net profit attributable to parents of 54.5 million yuan in the first half of the year, compared to a loss of 97.7 million yuan in the same period last year; the adjusted net profit attributable to parents reached 116 million yuan, improving from a loss of approximately 51 million yuan the previous year. Operational net cash flow also turned positive, reaching 142 million yuan. The gross profit margin increased to 67.7%, up 2.1 percentage points year-on-year, and revenue per capita grew 20.8% year-on-year, with expectations to surpass 700,000 yuan for the entire year. From these numbers, it is evident that Kingdee is witnessing two simultaneous phenomena: on one side, the subscription business from the past decade's cloud transformation continues to grow and starts to consistently release operational leverage; on the other hand, the AI-native business is growing at a rate far exceeding that of traditional SaaS and is beginning to make a real contribution to revenue and profit. The company's ability to turn losses into profits in the first half of the year is largely attributed to AI. According to management, about two-thirds of new subscriptions in the mid to large market chose AI-native products, indicating that AI is no longer an option but one of the core decision factors for new customers selecting Kingdee. With Kingdee's previous cloud transformation path continuing, the market previously recognized it as entering a harvest period, and by 2026, it is now seen that cloud business is beginning to serve as the infrastructure for AI commercialization. In other words, AI is not a random addition to Kingdee's existing business system. Now, with AI entering enterprise operational scenarios, it is essentially leveraging past accumulated business data, management processes, authority systems, and customer relationships. Without this foundational layer, AI can easily remain at the level of chat, Q&A, and office assistance; however, with ERP and SaaS as entry points, AI can further penetrate into core business processes of finance, procurement, supply chain, manufacturing, and human resources. The acceleration of this growth is fundamentally due to the significant value density created by AI for enterprises. Lingee: The Fulcrum for AI Implementation The AI-native products disclosed by Kingdee center around the "Lingee Enterprise Intelligent Agent Operating System," officially released on May 20, 2026. It is positioned as a unified entry point for enterprise-level AI, transforming the accumulated business data, management knowledge, organizational memory, and business rules into governable and reusable enterprise-level contexts. It is understood that Lingee's pricing model consists of "seat fees + usage points", where seat fees range from 58 yuan/year to 1,168 yuan/year. The usage includes both the token consumption of large models and the engineering capabilities Kingdee has built on top of those tokens. After its launch, Lingee quickly signed contracts with 26 clients and deployed over 40 intelligent agents across common fields such as finance, supply chain, procurement, and manufacturing. By early August, Lingee had expanded to over 100 clients; the company plans to onboard 10,000 clients onto Lingee in the second half of 2026, with approximately 80% expected to come from existing clients and around 20% from new clients. Kingdee's confidence in setting this goal comes from its extensive client base: over 40,000 active customers and over one million cumulative enterprise clients served. For these clients, Lingee represents an upgrade rather than a complete re-selection of suppliers atop their existing ERP and SaaS systems. Meanwhile, Kingdee has focused on deep AI integration in eight major industries: automotive and parts, equipment manufacturing, high-tech electronics, process manufacturing, life sciences, food and beverage, modern services, wholesale and retail. New clients include Hytera Communications Corporation, Boson Quantum, Shuanghui Group, Deer Siasun Robot & Automation, Kunlun Core, and Mianbi Intelligent, among others. Furthermore, efficiency improvements are already supported by concrete data: the expense audit intelligent agent has released 20% to 30% of manual auditing resources, the bank-enterprise reconciliation agent achieves 24/7 automated reconciliation, procurement payment audit agents have compressed the tri-matching process from hours to minutes, and the checkout agent has improved settlement efficiency by 30%. On the client side, Yike Group developed an equipment management system based on Lingee Build, increasing research and development efficiency by over 65% and shortening delivery cycles from weeks to days; Shuanghui Group's dealer sales order delivery time has been reduced by approximately 30%; and Wen's Group's development cycle has been cut by 50%. Consequently, Kingdee's market position in the AI field has received widespread recognition. The company is the only Chinese manufacturer included in Gartner's "Voice of the Customer for Product-Centric Cloud ERP" and leads globally in customer recommendation intent; it is also the sole Chinese entry in Gartner's "Magic Quadrant for PLM Software in Discrete Manufacturing." In multiple segments measured by IDC, Kingdee ranks first, with its public cloud revenue and market share for AI-enhanced enterprise-level ERP also ranking first, and Lingee has been included in Gartner's "Innovation Insights: Enterprise Intelligent Agent Capabilities" report. The Growth Curve is Becoming Steeper If the moat for traditional SaaS primarily comes from products and customer relationships, the moat in the AI era might further derive from "enterprise context." When AI begins to comprehend a company's business data, management rules, and long-term memory, deeply embedding within ERP, the cost of replacement naturally rises as usage time increases. Kingdee's chairman Xu Shaochun quoted "The Art of War" at the performance meeting, stating: "Victory comes from oneself or defeats the enemy. For Kingdee to do well, it is about enabling end-to-end business processes with AI, completing it through intelligent agents in a closed loop." This logic is being validated by data, with Kingdee presenting a set of comparative data in its performance PPT: the time taken for products to surpass 1 billion in revenue15 years during the software licensing era, 6 to 7 years for cloud SaaS subscriptions, while AI-native products are expected to just require 1 to 2 years. Based on current progress, AI-native revenue of nearly 300 million yuan has already been achieved in the first half of the year, with the second half being the traditional peak season, making the annual target of 1 billion yuan not overly ambitious. On the globalization front, the company has established a local service network in six markets: Malaysia, Thailand, Vietnam, Indonesia, Singapore, and Qatar, achieving direct delivery for tax compliance in these six countries, signing about 140 international business clients, with new clients including Yongqing Group, the Hong Kong Treasury, and CUCKOO International. During this period, the company made a total of approximately 775 million yuan in external investments, focusing on AI companies such as Moonlight Dark Side, Stepping Stars, and Silicon-Based Fluid, with a net cash reserve of about 3.5 billion yuan, having repurchased approximately 350 million Hong Kong dollars in shares through repurchase and employee equity platforms, canceling 1% of the total share capital. Of course, AI products are still in a phase of rapid iteration, and customer needs have not yet fully stabilized. Management acknowledges that, at this stage, Lingee's gross profit calculated by points may not necessarily exceed that of traditional products due to token costs; however, as model inference costs decrease and routing and platform engineering capabilities improve, the gross profit margin of AI products is expected to improve further. The company maintains its annual guidance: group revenue is expected to continue double-digit growth, AI-native product revenue exceeding 1 billion yuan, subscription revenue maintaining around 20% growth, and adjusted net profit margin rising from about 3.3% to around 7%, with operational net cash flow increasing by over 20% year-on-year. In the longer term, the company continues to pursue its 2030 goal of "reinventing Kingdee with AI," targeting a 50% revenue contribution each from AI-native and AI+SaaS. Currently, Kingdee stands at a junction with its established subscription model on one side and its accelerating AI-native business on the other, while an increasingly efficient organizational structure supports both. The synergy of these three aspects is pushing the profit ceiling of a mature SaaS company higher. The most significant change in Kingdee's mid-2026 performance lies not just in turning a profit from losses, nor in revenue growth of 13.6%, and certainly not solely in the 189% increase in AI-native revenuebut rather in the fact that AI-native product revenue has for the first time formed a nearly 300 million yuan scale, beginning to contribute significantly to Kingdee's revenue growth. Over the past decade, the clients, data, subscription models, and enterprise service capabilities accumulated by Kingdee during its cloud transformation have become the foundation for the next round of AI commercialization. Having taken ten years to transition from traditional software to cloud SaaS, Kingdee is now embedding AI back into the enterprise operational system in a shorter time. According to data from the first half of 2026, AI is not creating a new curve outside of the existing business, but rather reopening a larger enterprise AI management market atop the already successful SaaS business model Kingdee has established.