Kingsoft Cloud Q4 revenue reaches 2.76 billion yuan, setting a new record; adjusted operating profit achieves profitability again

Investing in stocks relies on Kingfinch Analyst Reports—authoritative, professional, timely, comprehensive—helping you uncover potential thematic opportunities!

On March 25, 2026, Kingsoft Cloud (NASDAQ: KC, HKEX: 3896) released its 2025 Q4 financial report. During the quarter, the company demonstrated a strong growth trend driven by intelligent computing cloud, with resilience in its core businesses, delivering a “double win” in both performance and the accumulation of business capabilities. The financial report shows that Kingsoft Cloud’s Q4 revenue reached RMB 2.76 billion, setting a new historical high, up 23.7% year over year; of this, public cloud revenue was RMB 1.90 billion, up 34.9% year over year; industry cloud revenue was RMB 860 million, up 18.4% quarter over quarter; adjusted gross margin was 17.1%, adjusted operating profit margin was 2.0%, and adjusted operating profit was RMB 54.586 million, achieving profitability again.

As the sole strategic cloud platform in the Xiaomi Kingsoft ecosystem, the company’s collaboration with ecosystem partners remains solid and continues to deepen. This quarter, revenue from the Xiaomi Kingsoft ecosystem reached RMB 800 million, up 63.2% year over year, accounting for 29.1% of total revenue. For full year 2025, the company’s related-party transactions with Xiaomi and Kingsoft approached the upper limit of the quota, reaching 94.3%.

Full-stack AI capabilities continue to evolve; intelligent computing cloud revenue reaches yet another new high

Facing the dual trends of rapid model iteration and large-scale AI application, Kingsoft Cloud continues to deepen technological upgrades and product iteration around foundational cloud and intelligent computing cloud, making thorough preparations for an explosion in AI inference demand. During the quarter, benefiting from simultaneous growth in both ecosystem and non-ecosystem businesses, the company’s intelligent computing cloud billings revenue reached RMB 930 million, up 95% year over year, accounting for 49% of public cloud revenue.

At the intelligent computing cloud layer, Kingsoft Cloud’s Xingliu platform has completed its strategic upgrade from a resource management platform to a one-stop AI training and inference full-process platform, with a leap in platform capabilities. Among these, as a full-chain cloud-native platform for robot development and deployment, Kingsoft Cloud Xingliu Robot Platform deeply integrates core links such as data collection, annotation, model development, training, deployment, and simulation, building a unified engine for data, models, and simulation tailored specifically for embodied scenarios; the Xingliu platform also launched the MCP Square, Prompt optimization, and AI search capabilities, helping enterprises realize agent development and deployment through platform-based approaches, and gradually building a new ecosystem centered on agent operations support.

At the foundational cloud layer, the company fully upgraded underlying technologies including computing, storage, and networking, achieving deep integration with AI application scenarios. During the quarter, based on its self-developed DPU architecture, the GPU elastic bare metal ebmp8v enables direct GPU computing power access and exclusive physical resources, while integrating cloud disk and high-performance file storage capabilities to serve AI training and high-performance computing scenarios; it also launched the ks3fs high-performance mounting tool, which fully complies with the POSIX standard, supports automatic two-way data synchronization, and enables convenient interoperability between object storage and local file systems.

Kingsoft Cloud Galaxy platform V3.3.0, newly released and covering the entire lifecycle of cloud construction, migration, use, and management, not only completes a comprehensive upgrade of intelligent computing capabilities, but also performs deep adaptation for full-stack localization, achieving breakthrough outcomes with “both better performance and lower costs.” The intelligent computing power scheduling platform V1.1, with its three-in-one architecture design of “integration—scheduling—operations,” promotes efficient utilization of computing resources and release of value, significantly improving computing resource utilization; the intelligent operations and maintenance platform GSO V3.3.0 focuses on AI intelligent computing scenarios, comprehensively enhancing AI infrastructure operations and maintenance capabilities, enabling a qualitative upgrade from traditional private cloud to AI intelligent computing cloud.

Scenario implementation accelerates breakthroughs; AI value moves deeper

AIGC is becoming an important carrier for AI empowering industries across the board, driving the transformation of AI from model capabilities to industry value in a comprehensive way. Based on proven successful experience, market reputation, and core solutions that can be reused, Kingsoft Cloud is converting full-stack AI capabilities into a key driving force for business growth.

Benefiting from the continuous expansion of the boundaries of its AI capabilities, the company not only firmly captures the diverse business needs within the Xiaomi Kingsoft ecosystem, but also expands business outside the ecosystem. This quarter, the company strongly supported a leading internet enterprise’s online multimodal model inference business, reached a substantive strategic cooperation with a unicorn autonomous driving company, and also provided AI service support for BanYu and ShaoNianDe, helping reduce costs and improve efficiency.

Using Beijing government cloud as a model, the company has achieved business expansion in Zhuhai, Shenzhen, Wuhan, Shanghai, Yunnan, and other locations. To date, Kingsoft Cloud has reached cooperation with more than 10 regions, jointly building government cloud and state-owned enterprise cloud projects, and working together to explore deep empowerment of large models for digital and intelligent transformation.

In the digital health sector, the specialty system and PMO service project of Yichang Central People’s Hospital, built under the “integrated development and governance” concept, has successfully completed acceptance. In addition, the company also released a healthcare intelligent operations management AI application based on the Data Agent architecture, upgrading hospital management from “post-event statistics” to “pre-forecasting and mid-course intervention,” significantly lowering the barriers for data application.

In the enterprise services sector, the company provided stable and efficient computing power support for a leading high-end manufacturing enterprise’s intelligent manufacturing, industrial vision, and AI R&D, achieving the key deployment of intelligent computing in core high-end manufacturing scenarios; at the same time, the green energy operations platform built for a clean energy service provider was successfully put into trial operation and went live, helping customers further deepen platform operations efficiency and drive large-scale deployment of business.

Kingsoft Cloud CEO Zou Tao said: “Looking ahead, we will continue along the path of high-quality, sustainable development. We firmly believe in and embrace AI strategic opportunities, move with the tide, and work with genuine effort—refining and accumulating capabilities in core product technology solutions, firmly seizing and deeply developing the vast business opportunities both within and outside the ecosystem, continuously improving profitability, and creating greater value for customers, shareholders, employees, and society.”

		Sina Disclaimer: Sina.com publishes this article for the purpose of conveying more information, and does not imply agreement with or verification of the viewpoints or descriptions. The content is for reference only and does not constitute investment advice. Investors act at their own risk.

A wealth of information, precise interpretation—available in the Sina Finance APP

Responsible editor: Song Yafang