Kling has recently made significant strides toward going public.

According to Bloomberg, citing sources with knowledge of the matter, Kuaishou’s video generation subsidiary, Kling AI, has engaged CICC, Goldman Sachs, and UBS to advise on a potential Hong Kong IPO. The company aims to debut as early as 2027, seeking to raise at least $1 billion. However, underwriting agreements are still in preliminary stages, with both the timeline and fundraising target subject to change.

In May, rumors surfaced regarding Kuaishou’s plans to spin off Kling. Two months later, Kuaishou announced a formal external financing round, allowing investors to subscribe up to approximately RMB 20.447 billion. This valued Kling at a pre-money figure of $15 billion and a post-money valuation of around $18 billion. Tencent, Alibaba, Baidu, along with several industry and financial investors, participated in the round.

By late August, new investors, including the National AI Industry Investment Fund, joined, fully subscribing to the offering. In late September, Kuaishou announced management changes, appointing Gai Kun as CEO of Kling AI, with Cheng Yixiao overseeing community science initiatives concurrently.

From securing financing to reorganizing its team and preparing for an IPO, Kling has been making a major capital move roughly every one to two months.

These developments are accompanied by an increasingly complex relationship between Kling and its parent company, Kuaishou.

Kling’s latest funding round gave it a post-money valuation of approximately $18 billion. In early October, Kuaishou’s market capitalization in Hong Kong even briefly dipped below this figure during certain trading sessions.

I. Kuaishou Nurtured Kling but Couldn’t Replicate Its Own Success

When Kling officially launched in June 2024, Kuaishou had a distinct advantage that other video model startups lacked: it was already a video-first company.

Kuaishou boasted a robust network of creators, short dramas, advertisers, and a content production, recommendation, and monetization system that processed vast amounts of video data daily. Under this framework, Kling would lower the barriers to video production, while Kuaishou would distribute these videos to hundreds of millions of users, monetizing them through advertising, e-commerce, and content consumption.

Together, model development, tools, distribution, and monetization formed a cohesive ecosystem.

In April 2025, Kling was elevated to a first-tier business unit, reporting directly to CEO Cheng Yixiao. Subsequently, Kuaishou continuously adjusted its model R&D and organizational resources, propelling Kling from consumer-grade video generation tools into film/TV, advertising, short drama, and enterprise API markets.

The results were striking. In Q1 2026, Kling’s revenue exceeded RMB 650 million, surging over 300% year-over-year (YoY); Q2 revenue surpassed RMB 850 million, up over 200% YoY. Half-year revenue totaled over RMB 1.5 billion, already exceeding the approximately RMB 1.1 billion for all of 2025.

In March, Kling’s annualized revenue run rate neared $500 million. For a product launched in 2024, this growth rate places it among the global elite in video generation companies.

In August, National Business Daily, citing internal Kling business documents, reported that overseas revenue accounted for nearly 80% of Kling’s Q1 2026 total, up from about 60% in Q1 2025. API revenue surpassed subscription revenue for the first time that quarter, reaching about 60%.

Among API revenues, so-called “super major clients” contributed over 80%, with eight of the top ten API revenue clients from January to May 2026 being overseas enterprises.

Thus, Kling’s monetization model has diverged from the initially envisioned integration within Kuaishou’s ecosystem. Instead, it increasingly resembles an independent supplier selling model capabilities to global video creators, AI application companies, and enterprise clients.

For example, a U.S. advertising production team could purchase Kling’s API and distribute generated videos on YouTube or Instagram. An overseas AI video product could integrate Kling in the backend, handling customer acquisition, billing, and user relationships independently.

However, while Kling earns from model API calls, Kuaishou may not necessarily capture corresponding advertising spending, user engagement, or e-commerce transactions.

Of course, for a Chinese company to bypass domestic internet platform traffic competition and directly monetize global professional clients represents a significant commercial breakthrough. However, synergy between this revenue and Kuaishou’s existing businesses cannot be assumed simply because they share a parent company.

ByteDance offers an alternative observable model.

Seedance powers Jiying, CapCut, and Jianying while providing model services to enterprises via Volcano Engine. ByteDance monetizes model API calls, retains users through creative tools, and handles subsequent distribution via content platforms.

Kling similarly has Kuaiying, Kuaishou’s main platform, and advertising businesses behind it. However, Kuaishou has not disclosed sufficiently comprehensive data to prove Kling’s overseas API clients have massively converted into revenue for other group businesses.

By June 2026, short drama supply on Kuaishou’s platform had grown over fivefold compared to January; in Q2, online marketing service spending driven by short dramas surged over 100% YoY.

Improved content production efficiency does not equate to platforms capturing all incremental value.

If an advertiser previously spent RMB 100,000 on ad materials but now only needs RMB 30,000, the saved RMB 70,000 could boost advertiser profits, increase ad budgets, or be captured by other platforms. Kuaishou can only prove this efficiency remains on its books through corresponding improvements in user engagement, ad spending, transaction volume, or profit margins.

Current Kuaishou financials offer no clear answer.

In Q2 2026, Kuaishou’s daily active users (DAUs) reached 412.3 million, up 0.8% YoY; online marketing revenue hit RMB 20.6 billion, up 4.4% YoY; live streaming revenue was RMB 8.7 billion, down 13.5% YoY.

What is certain is that Kling remains in a high-investment phase. According to Kuaishou’s July financing disclosure, assuming relevant business restructuring was complete, Kling’s 2025 revenue would be approximately RMB 1.1 billion with an unaudited net loss of about RMB 1.9 billion.

This means that while Kling has started generating real revenue, R&D, computing power, product, and market investments still far exceed current revenue. A video platform that took over a decade to accumulate cash flow now supports a business requiring continuous GPU purchases, new model training, and overseas client acquisition.

II. Kling Gains New Shareholders and a New Deadline

The July 2 financing announcement outlined a new capital structure for Kling: initial investors signed subscriptions worth approximately RMB 19 billion, with subsequent new investors filling the remaining quota for a total of about RMB 20.447 billion ($3 billion).

After fully utilizing the subscription ceiling and employee equity incentives, the Kuaishou ecosystem would still hold approximately 68.33% of Kling’s equity, maintaining control, with Kling’s operating results continuing to be consolidated into group financials.

Kuaishou did not fully divest Kling nor automatically shed its future losses through the spin-off.

Instead, it brought in another group of investors to share expansion costs while isolating Kling’s operations, management incentives, and capital market valuation.

Spin-offs themselves are not problematic.

The video large model industry where Kling operates is undergoing rapid technological iteration. Today’s model advantages could be reevaluated within months, making exit timelines critical for Kling’s new shareholders.

Financing terms stipulate that if Kling fails to complete an IPO meeting protocol requirements by October 30, 2031, or cannot fulfill specific restructuring items as agreed, investors may demand equity repurchases under certain conditions.

Repurchase pricing involves the original investment principal, simple interest returns calculated at 8% annually, and protocol-specified allocation adjustments.

This right does not unconditionally guarantee investor returns, nor does it directly impose full rigid payment obligations on Kuaishou’s parent company. According to disclosed agreements, the repurchase obligation subject and specific trigger conditions require differentiation.

However, it does impose a constraint on Kling that did not exist as an internal group business. Previously, as long as Kuaishou was willing to continue investing, Kling could adjust its pace according to group budgets. Now, with external capital introduced, the company must explain revenue growth, fund utilization, and listing progress to independent shareholders alongside technological competition.

The management team becomes crucial in this context.

On September 30, 2026, Gai Kun officially became CEO of Kling AI. Under the previously disclosed equity arrangement, he received approximately 3% equity with tenfold voting rights for certain shares, corresponding to about 23.62% of voting rights.

These special voting rights are tied to the CEO position and will expire upon departure; when transferring shares, the associated special voting rights also terminate. Kling has set a maximum 15% equity participation plan authorization quota for employee and core management incentives.

Kuaishou clearly hopes to retain talent capable of driving model R&D through independent company equity and governance mechanisms, having learned from past experiences. In August 2025, Zhang Di, Kling’s former core technical leader, left Kuaishou to join Alibaba. In August 2026, media reported that Kling technical backbones Wang Xintao and Wang Meng had also departed.

Zhang Di participated in Kling’s early technical system construction. After leaving Kuaishou, he joined an Alibaba team that launched the video generation model HappyHorse, entering the same market competition.

The departure of a key researcher may not immediately impact existing products but could introduce uncertainties for next-generation model development. Kling must continue releasing stronger models while binding core team interests to this pre-IPO company.

Meanwhile, Kling’s client structure exposes future revenue to competitive pressures.

According to business materials reported by National Business Daily, Kling’s API revenue accounted for about 60% of total revenue in Q1 2026, with over 80% coming from a few super major clients. Sources indicate that some overseas clients’ contributions in April were already affected by competition.

A professional video platform might choose Kling today for its generation quality, speed, and pricing. Next quarter, if ByteDance, Alibaba, MiniMax, or other model companies release superior products, clients will retest offerings.

For enterprises procuring model capabilities via API, switching suppliers does not necessarily require rebuilding entire products.

This competition differs from Kuaishou’s original short video business. Short video platforms rely on creator relationships, recommendation algorithms, content accumulation, and user habits; model APIs must repeatedly prove current capabilities and cost-effectiveness. Clients acquired last quarter may demand re-quoting next quarter.

Kling faces a competition that cannot be won through a single technological lead.

Epilogue

Kling indeed represents China’s fastest-commercializing AI video player.

It took just one year to scale annual recurring revenue (ARR) from $100 million to nearly $500 million; it is among the few Chinese AI applications generating substantial overseas revenue, achieving positive gross margins (excluding training costs); it has served as a technical partner at Cannes, with works entering Hollywood studios and domestic hit dramas.

However, nearly all these achievements were accomplished outside Kuaishou’s ecosystem. Over 70% of revenue comes from overseas, relying on $3 billion in external funding, independent entity incentives, relocating founders from the main platform for dedicated management, and a $18 billion primary market valuation—exceeding Kuaishou’s own approximately $16.3 billion total market cap.

Interestingly, while Kuaishou discusses “short-term pain”—a narrative repeated for five quarters during earnings calls—it has spent HKD 8.35 billion repurchasing 174.84 million shares under a HKD 16 billion plan announced in May 2024.

Repurchases support stock prices, while spin-offs unlock value. These simultaneous actions indicate Kuaishou’s management clearly understands why the market discounts its valuation.

Whether Kling can win this high-stakes gamble remains uncertain. After all, Kling 3.0 and Omni have a 15% option pool, endorsements from BAT and Middle Eastern capital, and five full years ahead.

*The featured image and illustrations in the text are sourced from the internet.