A：The highlights of Kingsoft's fourth quarter and full year 2025 earnings include achieving a quarterly revenue of RMB 2.76 billion, with a significant increase in intelligent computing services revenue. The AI business grew by 95% year over year, contributing 49% to public cloud services. The company also experienced robust ecosystem and external business segment growth, strong profitability with a 17.1% adjusted growth margin and a 2.0% adjusted operating margin, and an increase in non-ecosystem customers' revenue by 44% year over year.

A：The growth in public cloud services and enterprise cloud revenue signifies Kingsoft's solidified cooperation within the Xiaomi and Tsub ecosystem, capturing new external opportunities, and expanding its customer base beyond leading AI enterprises and internet giants. This growth is driven by the increasing demand for AI technology and the need for industrial intelligence solutions, signifying a trillion-dollar market opportunity and the technological leap across industries.

A：Kingsoft is leveraging AI in the healthcare sector through a platform that provides natural language insights into DRG cost control, enabling proactive hospital management and significantly reducing the barriers to data application. In the public service sector, Kingsoft is successfully entering key markets like Shanghai by utilizing its deep vertical expertise and enterprise service experience to capitalize on intelligent computing opportunities and generate synergies with its public cloud businesses.

A：The star flow platform has been recently upgraded with the introduction of features such as MCP (model context protocol), hub process optimization, and AI search. These features are designed to assist enterprises in developing and deploying AI agents through a unified platform, thus fostering an ecosystem centered around agent-based operations.

A：Looking ahead, the company plans to remain committed to its high-quality and sustainable development strategy. It will leverage the opportunities presented by the AI era, develop along with the industry, and refine its core technology. The aim is to capture market opportunities, enhance profitability, and create value for shareholders.

A：The company's revenue for the quarter was 2,761 million, representing a year-over-year growth rate of 24%. Specifically, revenue from public cloud services increased by 35% to 1,000.902 million from 1,000.410 million in the same quarter last year. Unprecedented explosive demand for AI business drove a 90% year-over-year revenue growth, amounting to 926 million.

A：Profitability saw substantial improvement driven by shifts in revenue structure, resulting in an adjusted growth margin that continued its upward trend. The adjusted EBITDA margin reached 78.5 million, up 12% points from 60% in the same quarter last year, though down from 30% in the last quarter. The year-over-year growth was fueled by a large contribution from AI-related business, which represents the primary cost component.

A：This quarter's total revenue was 2,000.761 million, with revenues from public cloud services accounting for 1,000.902 million, up 35% from 1,000.410 million in the same quarter last year. Real news from Enterprise Club services reached 850,000.9 million during this strong quarter, which was marked by a high volume of project competitions.

A：The company's investment in infrastructure to support intelligent cloud business growth led to a total cost of revenues of 2,000.296 million, an increase of 27% year over year. IDC costs rose by 30% year over year, solution development and service costs increased by 29% year over year, and depreciation and amortization costs grew by 48% year over year. These increases were attributed to investments in newly acquired and leased service and network equipment, especially for the AI business.

A：The adjusted gross margin for the quarter was 471 million, which increased 10% year over year and 20% sequentially. The improvement was primarily due to the expansion of revenue scale, the significant contribution from the AI business, and cost control measures. The Ras and service adjusted gross margin also increased, mainly due to the high contribution from enterprise clouds.

A：The adjusted operating expenses were 459 million, an increase of 45% year over year and 14% sequentially. The adjusted research and development expenses were 181 million, up 7% from the same quarter last year, while adjusted sales and market expenses were 111 million, a 3% increase year over year. Adjusted general and administrative expenses fell by 25% year over year. The adjusted operating profit was 55 million, a significant increase from 24 million in the same period last year, primarily due to revenue scale expansion, gross profit growth, and expense control. The non-GAAP EBITDA margin also increased to 28%, from 60% in the same quarter last year, mainly reflecting the company's strong commitment to AI and cloud computing development, strategic business structure adjustments, and cost control.

A：The company has launched the Mimo V2 series models, which received positive market feedback. The role and positioning within the X and King service strategy involve forming a team or portfolio of solutions with Kingsoft staying disciplined in not developing large language models, allowing Xiaomi to develop newer models. The company's strategy is to monetize service capabilities both in training and inference areas.

A：The company acknowledges the significant pricing increase from the supply chain side last year and strategically stocked up some key components in anticipation. Regarding current prices, they intend to not increase pricing for existing contracts with underlying resource stocking. However, for new customers and significant usage increases, there will be price hikes. The company aims to pass through some upstream cost increases to customers and adjust prices based on demand.

A：The company sticks to two principles: not increasing prices for existing contracts with stocked resources and allowing price hikes for new customers with significant usage increases. They aim to pass through some upstream cost increases to customers and increase prices based on demand to improve profitability.

A：The shift to a prediction and subscription model is seen as an inevitable stage of AI and large language model development. The company has launched a stock-flow platform, a neutral platform hosting open source models including those from Xiao, for model as a service business, which is the company's fastest-growing business. They provide both training and inference services to customers.

A：The company has not noticed Volcano Engine's price change but believes that under the current market dynamics of explosive demand growth and high supply chain costs, such low pricing in the real world would not be applicable. They focus on price hikes from other players like Alibaba, and the increase in catalog price does not necessarily reflect actual prices companies engage in.

A：For 2026, the company expects total CapEx and controlled assets to exceed Mv ed billion, with approximately half allocated for data centers. They plan to access more assets through short and long-term leases to minimize upfront capital expenditure. There are no equity finance plans for 2026, and capital expenditure will come from financing, strategic customer pre-pay, and commitment credit facilities. The company expects a growth rate acceleration and an improved EBITDA rate in 2026.

A：The top five non-ecosystem customers have shown strong year-over-year revenue growth, with more than half the potential demand coming in for inference versus training. The company is seeing extremely large demand from outside the ecosystem, which is higher than the demand from the ecosystem itself. The revenue from these demands will depend on the resources secured and delivered to customers. The stock-flow platform is growing fast with a better profit margin due to increasing agent applications.