Kingsoft Cloud’s Q2 2026 Financials: AI Cloud Drives 82% Growth

A financial report is more than just numbers on a page. It’s a story. The raw data tells you what happened, but the real journalism lies in deciphering the why and the so what. The latest earnings release from Kingsoft Cloud, a major Chinese cloud service provider, is a textbook case of a company in the midst of a profound and costly strategic pivot. The headline figures are undeniably strong: record revenue, a stunning 82% surge in AI cloud billings, and the company’s first-ever positive GAAP operating margin. Yet, beneath this surface of robust growth lies a more complex and capital-intensive narrative, one that speaks volumes about the global scramble for AI supremacy and the immense price of admission.

Let’s start with the undeniable good news. For the quarter ending June 30, 2026, Kingsoft Cloud reported total revenues of RMB 3.07 billion, a 30.8% jump from a year ago. The engine of this growth is unmistakable. Their public cloud segment, now 77% of total revenue, soared by 45.1% year-over-year. Driving that surge is what the company calls its “AI cloud business,” where gross billings exploded by 82%. This isn’t just incremental growth; it’s a fundamental reorientation of the company. As CEO Tao Zou noted, AI cloud now represents 56% of public cloud revenue. In the span of a few quarters, Kingsoft Cloud has effectively rebranded itself from a general-purpose cloud provider to an AI infrastructure specialist.

This strategic shift is finally translating into the metric that has long eluded many cloud players outside the industry’s apex: profitability. The company recorded an operating profit of RMB 23 million under standard accounting rules (GAAP), a landmark achievement compared to a loss of RMB 327 million a year prior. Their adjusted operating profit margin hit 4%. For CFO Yi Li, this improvement is directly attributed to “AI demand tailwinds and operational optimization.” The message to Wall Street and investors in Hong Kong is clear: the heavy investment phase is beginning to pay off.

However, this is where a seasoned analyst must look beyond the profit line and into the cash flow statement and balance sheet. The path to AI-driven profitability is paved with extraordinary capital expenditure. In the second quarter alone, Kingsoft Cloud spent RMB 3.3 billion on capital expenditures, including assets acquired through leases. That’s a staggering sum, roughly equivalent to 107% of the quarter’s total revenue. This spending is reflected in the soaring cost of revenues, which jumped nearly 30% year-over-year. A deep dive into those costs is revealing.

• Depreciation and amortization expenses nearly doubled, skyrocketing from RMB 552 million to RMB 963.8 million.
• The company explicitly states this is “mainly due to the depreciation of newly acquired and leased servers, and network equipment which were mainly related to AI cloud business.”
• IDC costs—what they pay for data center space and power—rose by 23%.
• This is the physical, gritty reality of the AI boom.
• It’s not just software; it’s about building and powering colossal, energy-hungry data halls filled with expensive, rapidly-depreciating hardware.
• The financial strain of this build-out is evident on the balance sheet.

The company’s cash and cash equivalents fell by over RMB 1.3 billion in just six months, which management directly links to “the investment into the procurement of computing power equipment.” This creates a fascinating, high-stakes financial dynamic. On one hand, the AI business is growing at a blistering pace and improving overall margins. On the other, it is consuming cash at a voracious rate to feed that growth. The company’s enterprise cloud business, its more traditional offering, saw revenues essentially flatline, decreasing 1.3% year-over-year. This suggests resources and focus are being decisively allocated toward the AI future, potentially at the expense of the existing, stable business. It’s a calculated bet that the AI market’s growth will outpace the immense capital burn required to serve it.

The use of non-GAAP metrics here, as with many tech firms, is crucial for understanding management’s perspective. Their non-GAAP EBITDA, which adds back huge expenses like depreciation, reached RMB 1.1 billion, with a remarkable margin of 35.8%. This metric is favored by management because it shows the underlying operational cash generation potential before the accounting weight of their massive hardware investments hits the income statement. It’s a way of saying, “Look at the powerful business model, once you ignore the cost of building the factory.” But investors cannot ignore that cost. The GAAP net loss, though narrowed significantly, still stood at RMB 93 million.

From my vantage point in the Financial District, this report encapsulates the defining challenge for second-tier cloud providers globally. The AI race is not optional; it’s existential. Kingsoft Cloud is demonstrating that it can capture this demand and even achieve a form of operating leverage. However, the financial statements scream that this is a scale game of terrifying proportions. The company is betting its future on its ability to continuously invest billions into computing power, hoping that AI revenue growth will eventually outrun the depreciation curves and interest expenses on that capital. It’s a bet on a future where AI is not a niche but the primary utility of the cloud. The Q2 2026 results show they are winning battles on growth and marginal profitability. The war for sustainable, cash-flow-positive leadership in the AI cloud, however, remains a fiercely expensive and open campaign.