Kuaishou's AI Spinoff Raises Questions as Unit Nears Two-Thirds of Parent's Value
Kuaishou Technology is engineering a spinoff of its AI video-generation unit Kling AI at a $20 billion target valuation, a move that has exposed a structural paradox: a subsidiary generating less than 1% of group revenue is being priced at roughly two-thirds of its parent company's entire market capitalization.
Shares of Kuaishou surged more than 11% at the open on May 12, 2026, briefly touching HK$57.40 intraday after reports emerged that the company was in discussions with investors — including Tencent — to raise $2 billion at that valuation. Kuaishou confirmed in a midday filing that its board is evaluating a restructuring of Kling AI's assets and business, which may involve introducing external financing. By the close, the stock's total market capitalization stood at approximately HK$228.7 billion (roughly US$29.2 billion at prevailing exchange rates), making Kling AI's proposed valuation equivalent to 68% of the entire listed entity.
That arithmetic is not a rounding error. It is a market verdict.
Kling AI's Metrics Rewrite the Valuation Playbook
As of end-April 2026, Kling AI's annualized recurring revenue (ARR) reached $500 million—doubling from pre-Chinese New Year levels, according to reporting by LatePost. The platform has accumulated more than 60 million global registered users, 12 million monthly active users, over 30,000 API enterprise clients, and a peak daily video generation volume exceeding 5 million clips, with cumulative output surpassing 600 million videos.
Yet in Kuaishou's Q4 2025 earnings, Kling AI contributed RMB 340 million (US$47.2 million) in quarterly revenue against group total revenue of RMB 39.6 billion (US$5.5 billion) — a 0.86% share. Few comparable valuation gaps between parent and subsidiary businesses have emerged in China's internet sector in recent years. Baidu's spin of Kunlun Chip and ByteDance's partial separation of Volcano Engine never produced a subsidiary-to-parent valuation ratio approaching this level.
The market is not pricing Kling AI on what it earns today. It is pricing what Kuaishou's core short-video business is no longer worth.
Sora's Collapse Validates Kling AI's Commercial Architecture
To understand why a $20 billion valuation is even being discussed, investors point to what happened to OpenAI's Sora. From its early 2024 debut through the launch of a standalone application in late September 2025 and its subsequent shutdown on March 24, 2026, Sora generated an estimated US$2.1 million in cumulative revenue, according to mobile-data firm Appfigures—a figure that underscores the chasm between technical spectacle and commercial durability.
Kling AI took the opposite trajectory. Embedded within Kuaishou's short-video ecosystem, it inherited a data flywheel of hundreds of millions of clips and a captive base of content creators facing genuine production pressure. The commercial use cases that sustain its ARR—AI-generated short dramas, advertising creatives, game cinematics, film pre-visualization, and e-commerce video—demonstrate meaningful repeat-purchase dynamics. Sora lacked all of this. The comparison is not about model quality; it is about whether a closed-loop business model exists.
Sora's exit has created a global market vacuum in premium AI video generation. Kling AI, with a verified commercial model and the largest non-Anglophone user base in the category, is positioned as the primary beneficiary.
ByteDance's Shadow Looms Over Every Kling AI Forecast
The competitive threat that actually determines Kling AI's long-term trajectory is not a Western rival—it is ByteDance's Seedance. At the 2026 Chinese New Year CCTV Gala, Seedance 2.0 generated nationally televised footage of ink-wash galloping horses and month-by-month floral sequences that prompted industry figures, including Black Myth producer Feng Ji, to call it the world's strongest AI video tool without qualification.
ByteDance's structural advantages over Kuaishou are difficult to ignore. Kuaishou's full-year 2025 revenue was RMB 142.8 billion (US$19.8 billion); ByteDance's is estimated at several multiples of that figure. ByteDance also controls TikTok and Douyin as distribution channels, Volcano Engine as cloud infrastructure, and CapCut as a video editing tool with natural integration points for generative video. Every resource dimension—compute budget, talent density, traffic scale—favors ByteDance.
The spinoff logic, viewed through this lens, is defensive capital strategy. A standalone Kling AI entity with $2 billion in fresh capital and Tencent as a strategic backer creates an independent war chest and a political counterweight against ByteDance's resource dominance. It mirrors the fundraising logic of China's large language model startups: without a dedicated capital channel, a well-resourced competitor can simply outlast you.
Spinoff Structure Risks Hollowing Out Kuaishou's Core Identity
The transaction's strategic benefit to Kuaishou is real but carries a structural cost that investors should not underestimate. Kuaishou's core business—a short-video platform with hundreds of millions of daily active users, an e-commerce GMV that crossed the RMB 1 trillion threshold in 2025, and a live-streaming monetization engine—has seen user growth plateau, advertising revenue decelerate, and live-gifting revenue face regulatory and behavioral headwinds. Capital markets have consistently valued Kuaishou as a discounted proxy for ByteDance's Douyin, chronically one step behind and one tier smaller.
Kling AI, elevated to a first-tier business unit reporting directly to CEO Cheng Yixiao in April 2025, has become Kuaishou's sole credible re-rating narrative. The spinoff is not primarily about unlocking subsidiary value—it is about using Kling AI's AI-era multiple to defend Kuaishou's aggregate market capitalization.
The risk calculus runs in both directions. If Kling AI lists successfully at $20 billion, Kuaishou's retained equity stake generates paper wealth that dwarfs Kling AI's current revenue contribution. But as Kling AI's independent identity strengthens, Kuaishou's residual business risks being re-rated as a pipeline asset—a holding company whose operating core generates cash but commands no growth premium. The brighter Kling AI shines, the more Kuaishou's short-video platform looks like its vessel rather than its engine.
AI Video Generation Sector Approaches Its Valuation Stress Test
The sector-level question is whether the market is pricing a durable industry or a concentrated early-adoption surge. DataEye estimates China's AI short-drama market at RMB 16.8 billion (US$2.3 billion) in 2025, projected to reach RMB 24.4 billion (US$3.4 billion) in 2026. Kling AI's standalone valuation at $20 billion is approximately six times the projected scale of its single largest commercial use case.
The structural difference between AI video generation and text-based AI compounds this concern. Subscription products like ChatGPT generate daily engagement and high user lifetime value. AI video generation is project-based consumption: users generate a batch of clips, exit, and return only when a new project demands it. That pattern resembles software tool purchasing more than platform subscription, implying structurally lower user lifetime value than investors may expect from text-based AI comparables.
Kling AI's ARR growth from near-zero to US$500 million in under two years is genuinely exceptional. The open question — and the central uncertainty embedded in a $20 billion valuation — is how much of that growth reflects durable B2B repeat purchasing versus early-adopter experimentation and subsidized pricing. Sora's lifecycle proved that impressive growth velocity and terminal failure are not mutually exclusive in this category.
Related Coverage:
Kuaishou's Kling AI Targets $20B Valuation in Planned Spinoff