Kingsoft Cloud (KC) Report Interpretation

Overview

This conference-takeaways report covers Kingsoft Cloud's AI-led cloud growth, capex plan and path to profitability. Goldman Sachs views KC as a beneficiary of Xiaomi's AI investment and broader demand from AI and internet customers, while retaining its Buy rating and US$18 target price.

Core views

The central growth driver is AI cloud. In 2Q26, total revenue grew 31% year-on-year, led by AI cloud revenue growth of 82% to Rmb1.3bn. AI cloud accounted for 56% of public-cloud revenue and 43% of total revenue. Within the business, MaaS generated Rmb120mn, or 9% of AI-cloud revenue. By contrast, enterprise-cloud revenue declined 1% year-on-year, partly because an accounting change moved some recurring revenue into public cloud. Xiaomi and Kingsoft revenue rose 28% year-on-year and represented 26% of revenue; management expects their growth to accelerate in 3Q26. Revenue from the top five non-ecosystem customers increased 51%, spanning AI labs, internet companies, embodied AI, smart-driving and fintech customers. MaaS is presented as an important higher-margin element of the AI-cloud mix. KC provides 120 models to 230 customers, mainly pan-internet key accounts, with use cases split between AI coding (80%) and AI workflow (20%). The report distinguishes lower-margin API resale of closed-source models, where gross margin is 10%, from higher-margin token sales through hosting open models, which do not require revenue sharing. Management also sees MaaS as offering greater flexibility to procure computing power and considers KC's inference-optimization capability broadly close to that of AI labs. Margin progression depends on both mix and contract economics. Management expects gross profit margin to stabilize or improve moderately versus 2Q26. MaaS has the highest gross margin despite revenue volatility; traditional public cloud follows, aided by lower depreciation and rising prices; and AI computing is lower-margin because it largely uses long-term key-account contracts, although small customers could theoretically generate 40–50% gross margin. CDN remains the lowest-margin segment. The key variables are the mix of key accounts versus SMEs, contract length and customer prepayments, which can cover 70–80% of direct capex. Management expects 2026 EBITDA margin of about 35%, supported by higher AI-cloud EBITDA margins. GAAP EBIT turned positive for the first time in 2Q26, helped by higher gross margin, operating-expense discipline and credit-loss reversal, and management aims to keep it positive; it also targets positive adjusted net profit in 2027. The four- to five-year depreciation policy, with no residual value, means depreciation from the first AI-capex batch in 2H23 should decline notably from 2H27. The investment case also rests on management's planned capacity expansion and its asserted payback. KC's base-case 2026 capex budget is Rmb15bn, including direct procurement and computing-power rentals, compared with Rmb6.2bn in 1H26 and Rmb8.2bn and more than Rmb4bn in 2025 and 2024, respectively. After price increases in 1H26, computing-power prices have become more stable, which management says has improved customer commitment. AI IaaS payback is estimated at 3.5 years on a revenue basis and shorter on a cash-flow basis when customer prepayments and leased computing power are considered. With China server costs rising, KC is prioritizing AI-training demand from AI labs and cloud-service providers over MaaS and is exploring lower-cost supply in ASEAN, despite higher land and power costs, in response to ecosystem-customer demand. For enterprise cloud, management believes traditional customers are reallocating IT budgets toward AI. KC sees potential to capture incremental workflow budgets tied to labor replacement, though the opportunity is still early-stage. It cites hybrid-deployment capability as an advantage, while noting that sales and deployment talent is critical to execution. Goldman Sachs characterizes KC as a leading Chinese cloud-service provider with the highest AI revenue contribution among China cloud peers at 31% in 2025. It highlights visibility from Xiaomi, for which it forecasts 46% revenue CAGR in 2025–28E, and views KC as a beneficiary of Xiaomi's increased AI investment and physical-ecosystem integration. The report also points to continuing AI-training demand from other internet and AI companies. Goldman Sachs is Buy rated and sets a 12-month US$18 target price using DCF with a 10.3% WACC and 3% terminal growth rate.

Analysis framework

The report combines management commentary from the Asia Leaders Conference with 2Q26 operating data, segment revenue and margin analysis, customer and contract-mix discussion, and capex-payback assumptions. Goldman Sachs then supports its rating with a DCF valuation using stated discount-rate and terminal-growth assumptions.

Methodology notes

• Valuation methodsDCF (Discounted Cash Flow)

  Discounted cash flow valuation

  Goldman Sachs derives its US$18 12-month target price by discounting projected cash flows using a 10.3% WACC and a 3% terminal growth rate.

• Industry AnalysisVolume-price decomposition

  Segment and customer revenue-mix analysis

  The report separates AI cloud, MaaS, enterprise cloud, ecosystem customers and non-ecosystem customers to explain the sources of growth and the implications for gross margin.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Kingsoft Cloud (KC)

  Primary covered cloud-service provider and reported beneficiary of AI-cloud demand, Xiaomi investment and broader AI-training demand.

  Strengths

  AI cloud grew 82% year-on-year in 2Q26; MaaS carries the highest gross margin; management cites hybrid-deployment capability and customer-prepayment support for capex.

  Weaknesses

  Enterprise-cloud revenue declined 1% year-on-year in 2Q26, partly due to an accounting reclassification; CDN carries the lowest gross margin.

  Comparison

  Goldman Sachs states KC had the highest AI contribution to revenue among China cloud-service providers, at 31% in 2025.

  Risks

  Supply-chain and high-end-chip availability, peer competition, lower AI investment by key customers, capex funding constraints and dilution.

Key data

• 2Q26 total revenue growth+31% yoyGrowth was led by AI cloud.
• 2Q26 AI cloud revenueRmb1.3bn; +82% yoyAI cloud represented 56% of public-cloud revenue and 43% of total revenue.
• 2Q26 MaaS revenueRmb120mnEquivalent to 9% of AI-cloud revenue.
• Xiaomi and Kingsoft revenue growth+28% yoyThe segment represented 26% of 2Q26 revenue; management expects faster growth in 3Q26.
• Top five non-ecosystem revenue growth+51% yoyCustomers include AI labs, internet companies, embodied AI, smart driving and fintech companies.
• 2026 EBITDA margin targetc.35%Management attributes the target to higher AI-cloud EBITDA margin.
• 2026 capex budgetRmb15bnIncludes direct procurement and computing-power rental; versus Rmb6.2bn in 1H26, Rmb8.2bn in 2025 and more than Rmb4bn in 2024.
• AI IaaS revenue payback period3.5 yearsManagement says cash-flow payback is shorter with customer prepayments and leased computing power.
• Goldman Sachs revenue forecastRmb9,558.6mn in 2025; Rmb12,816.4mn in 2026E; Rmb16,428.7mn in 2027E; Rmb19,988.6mn in 2028EForecast revenue trajectory shown in the report.
• Goldman Sachs EBITDA forecastRmb2,328.1mn in 2025; Rmb4,541.0mn in 2026E; Rmb6,443.4mn in 2027E; Rmb8,367.7mn in 2028EForecast EBITDA trajectory shown in the report.

Impact & implications

The report argues that AI cloud can increase KC's revenue growth and support profitability as higher-margin MaaS and public-cloud offerings expand, while prepayments and rentals improve the economics of capacity investment. It identifies Xiaomi's AI investment and external AI-training demand as the main sources of revenue visibility, but execution depends on capacity, customer mix and deployment capability.

Risks

• Supply-chain disruption or an inability to secure high-end chips could constrain expansion.
• Heightened competitive pressure from peers could weaken the outlook.
• AI investment by key customers, including Xiaomi, could be below expectations.
• KC may be unable to secure sufficient funding for capex investment.
• Dilutive financing activities could pressure shareholders.

What to watch

• Whether AI cloud continues to drive growth and whether Xiaomi and Kingsoft revenue growth accelerates in 3Q26.
• Progress in maintaining or improving gross profit margin, reaching about 35% EBITDA margin in 2026, and sustaining positive GAAP EBIT.
• Execution of the Rmb15bn 2026 capex plan, customer prepayments and the stated AI IaaS payback period.
• Demand from AI labs, cloud-service providers and other internet and AI customers, along with computing-power availability and costs.