NUPIAO
On the evening of July 2, Kuaishou Technology dropped a major announcement on the Hong Kong Stock Exchange: its independent operating entity, Beijing Kling, just wrapped up an external capital increase of up to $3 billion, giving it a post-money valuation of roughly $18 billion. This instantly made Kling AI the biggest single fundraising round ever for a video generation model company, officially kicking off its standalone commercial journey.
The round was co-led by a heavyweight lineup including CPE Yuanfeng, GuoFang Venture Capital, BlueFive, Tencent, Zhongguancun Science City Fund (together with Uni-Science Investment), and CITIC Securities, alongside dozens of tier-one institutions. On top of that, industrial players like Alibaba Cloud and Baidu, plus entertainment industry investors such as Huace Film & TV and Mango Industry Investors (Houwei Capital), jumped in with gusto.
According to the announcement, the initial 21 institutions together poured in RMB 13.8236 billion, while 15 additional investors signed up to add another RMB 5.2235 billion, pushing the total capital increase ceiling to RMB 20.4471 billion—representing 16.67% of Kling’s enlarged registered capital. Every single penny will be channeled into model iteration, commercial expansion, and building a truly global team.
The deal also activated three long-term equity incentive plans, with a total authorized quota covering 15% of the post-enlargement registered capital and a uniform 10-year validity period. These plans span equity awards, limited partnership indirect shareholding, and stock options, carefully tailored to executives, core R&D staff, and external technical advisors.
The first batch granted 7.1651 million equity and option units, accounting for 7.45% of the enlarged total shares. Among the highlights, Kuaishou Chairman and Kling Chairman Cheng Yixiao got 1% equity at zero cost, locked in for three years with a six-year departure clawback. Kling CEO Gai Kun scored a 3% stake and also enjoys super voting rights of up to 4% of the share capital with ten times the voting power; however, if he leaves or transfers his shares, those differential voting rights automatically fizzle out.
If the funding cap and all equity incentives are fully realized, Kuaishou—through entities like Lucky Labs—will hold a solid 68.33% of Kling, keeping the financial consolidation intact. That means all of Kling’s revenue and losses will continue flowing straight into Kuaishou Group’s financial reports. External investors will collectively hold 16.67%, and the employee incentive pool will account for the remaining 15%.
The announcement also spells out a five-year non-compete promise: once the restructuring is done, until Kuaishou no longer controls Kling or five years have passed (whichever is later), the group won’t directly or indirectly control any entity that primarily dabbles in the video generation model business.
The companion asset restructuring plan was locked in at the same time. Kuaishou’s scattered Kling R&D and operation subsidiaries will be fully integrated within nine months after the final payment of the capital increase. Non-AI video businesses will be carved out and handed back to the parent, while all models, qualifications, and overseas operating entities will be unified under Beijing Kling. Simultaneously, the company is pushing through AI algorithm filing, value-added telecom qualifications, and the full suite of overseas investment ODI approvals.
All external investors have a share redemption right: if Kling fails to pull off an IPO by October 30, 2031, or if the restructuring, qualification filing, and overseas entity acquisition aren’t completed on schedule, investors can demand redemption at the original principal plus 8% annual simple interest. Kling Group will bear joint liability for that.
Kling AI (brand name Kling AI) is Kuaishou’s self-developed multimodal video generation model system, steered by Tsinghua PhD Gai Kun.
Public records show Gai Kun has a dual AI R&D background spanning both Alibaba and Kuaishou. He joined Kuaishou in 2020, overseeing community recommendations and visual model development. When OpenAI dropped Sora demo videos in 2024 but didn’t open it up to the public, Gai Kun quickly figured that overseas companies would struggle to roll out commercial products anytime soon. So he rallied his team and accelerated Kling 1.0, aiming to build the world’s first industrial-grade, photorealistic video generation model that ordinary users could actually use right away.
In April last year, Kuaishou made an internal organizational leap and officially elevated Kling to a first-level independent business unit, reporting directly to Kuaishou founder Cheng Yixiao. Remarkably, this was the only new first-level business unit Kuaishou created in nearly three years. Then in August, Gai Kun took on the mantle of Kling’s technical lead, unifying product, algorithm, and R&D from end to end, and assembling a dedicated AI video R&D army of over a thousand people.
Kuaishou’s financial data reveals that Kling’s full-year revenue in 2025 was about RMB 1.1 billion, with a net loss of RMB 1.9 billion. By March 2026, its annualized revenue run rate had already hit $500 million, with paid members and enterprise API customers showing rock-solid retention. The three core revenue engines—e-commerce, short dramas, and film/TV advertising—are truly firing on all cylinders.
According to Kling AI’s own operational data, as of June this year, global users had smashed through the 100 million mark—a 67% jump from the 60 million at the end of 2025—covering 224 countries and regions. Enterprise customers now stand at nearly 50,000, also up roughly 67% from 30,000 at the end of 2025.
On the commercialization side, Kuaishou’s Q1 2026 financial report showed Kling AI’s quarterly revenue surpassing RMB 650 million, a year-on-year surge of over 300%. In March 2026, ARR was tantalizingly close to $500 million, ballooning nearly 400% within a single year. In just two years since launch, Kling AI evolved from version 1.0 to 3.0 Omni, completing 26 iterations in the past year alone, while churning out 122 published papers and 21 open-source projects.
When explaining the rationale behind the restructuring, Kuaishou pointed out in its latest announcement that, on one hand, it would better reflect the true value of Beijing Kling and the Kling AI-related assets and businesses within Kuaishou Group. On the other hand, Beijing Kling’s business will naturally attract a different breed of investor—those laser-focused on AI video generation development and application—which is a far cry from Kuaishou Group’s own content community and social platform business model.
Furthermore, the move is set to boost Beijing Kling’s image among customers, suppliers, and potential strategic partners, strengthening its hand in negotiations and opening doors to more business. At the same time, it allows Kuaishou Group (excluding Beijing Kling) to allocate its financial resources far more efficiently.
A report from Intel Market Research underlines that the global market for pure text-to-video and image-to-video generation was a modest $1.23 billion in 2025, but the compound annual growth rate from 2026 to 2034 is projected at a scorching 46%, with the market size expected to smash through $21.61 billion by 2034. That makes it one of the fastest-growing segments in the entire AIGC universe.
Meanwhile, data from Fortune Business Insights highlights that North America has long gobbled up over 40% of the global AI video market, but China’s slice in 2026 is forecast to reach $49 million, putting it in the lead across the Asia-Pacific region.
In the domestic AI video arena, two distinct paths have emerged: mass-market traffic tools and industrial-grade commercial models. For instance, ByteDance’s Jimeng AI leans heavily on the colossal traffic of Douyin and CapCut, focusing on lightweight, free editing tools, whereas Kling is squarely aimed at enterprise-level industrial batch production scenarios.
Some industry analysts reckon that this monster funding round for Kling will completely redraw the competitive landscape. Small and medium-sized video AI startups are about to face a triple squeeze from computing power, capital, and customer resources, and a wave of industry mergers and acquisitions could well hit in the next two years.
