Author | Wen Yehao

Editor | Hao Xin

On May 27, Kuaishou unveiled its financial report for the first quarter of 2026.

The report disclosed that Kuaishou's revenue for the quarter hit RMB 33.716 billion, marking a 3.4% year-on-year increase. Under non-IFRS standards, the adjusted net profit stood at RMB 3.374 billion, a 26.3% decline from the previous year.

For Kuaishou, the current landscape is somewhat intricate.

While revenue continues to climb and user numbers expand, indicating a robust core business, the market's focus has shifted. Previously, Kuaishou's daily active users (DAUs), monthly active users (MAUs), and business segments like e-commerce, advertising, and live streaming were under the spotlight. Now, as the internet realm pivots towards AI, Kuaishou's potential hinges more on Kling.

The financial report revealed that in the first quarter of 2026, Kling AI delivered remarkable results, generating over RMB 650 million in revenue, a staggering 300% year-on-year surge. Coupled with recent rumors of a spin-off, this has positioned the short-video giant at a pivotal juncture.

Kuaishou's Future Lies with Kling

For a seasoned internet player like Kuaishou, the real concern isn't slowing growth but a lack of innovation amidst ongoing expansion.

In recent years, Kuaishou has teetered on the edge of this phase.

Over the past few years, Kuaishou has ventured into diverse territories, from aggressive e-commerce expansion to diversifying into local services, always in search of new narratives. While there have been successes and failures, most endeavors merely extended Kuaishou's core business without significantly altering its overall trajectory.

Take GMV, for instance. Five years ago, GMV was a key metric used to justify Kuaishou's "Old railway economy" (Iron Family Economy) during its Hong Kong Stock Exchange listing. That year, Kuaishou's e-commerce GMV soared by 78%, showcasing immense potential.

Five years on, e-commerce has become deeply ingrained in Kuaishou, transforming from a "darling" to a "mainstay," with annual GMV growth slowing to 12.9% last quarter. Notably, starting this year, Kuaishou will no longer disclose GMV separately.

While its commercial value persists, its narrative allure has waned.

From a user perspective, the financial report showed that in the first quarter of 2026, Kuaishou's average DAUs reached 412.7 million, a slight year-on-year increase, while average MAUs hit a record high of 771.7 million.

During its rivalry with Douyin, reaching a new MAU high would have been cause for celebration. However, this key operational metric now garners little attention.

Kuaishou's new narrative, naturally, revolves around AI.

Currently, many players are embracing AI, even hastily rebranding themselves as "AI companies." However, most merely amplify AI's narrative appeal, with stories often outweighing practical implementation.

While Kuaishou also weaves AI tales, it can be considered a pragmatic player.

In recent years, AI has been seamlessly integrated into Kuaishou's recommendation systems, advertising, content generation, and merchant services, playing a pivotal role in enhancing efficiency across the entire content ecosystem.

However, efficiency gains are subject to diminishing returns. While early results were immediately noticeable, once the platform fully absorbs AI, it downgrades from a novel variable to a normalized operational tool—important but lacking the initial allure, with limited narrative longevity.

In other words, while Kuaishou has strived to become an AI company, the market seems to only recognize Kling.

This discrepancy is also evident in recent news about Kuaishou's potential spin-off of Kling.

According to previous media reports, Kuaishou is considering spinning off Kling AI and initiating financing at a valuation of $20 billion. Kuaishou's subsequent announcement confirmed that the board is evaluating a restructuring plan for Kling AI's assets and businesses, which may involve external financing but remains in the preliminary stages, with no final agreements signed yet.

The market is abuzz about the valuation gap between Kling and Kuaishou.

As of now, Kuaishou's market capitalization in Hong Kong stands at around $25 billion. If Kling, accounting for less than 2% of revenue, is valued at $20 billion, only $5 billion would remain for Kuaishou. Ultimately, the capital market's willingness to price a new business still in its high-growth investment phase at nearly 80% of Kuaishou's market cap underscores the allure of a compelling AI narrative.

However, if the goal were merely to tell a story, Kuaishou wouldn't need to rush into spinning off Kling. Keeping it within the "original family" would suffice, with a few mentions in financial reports to stir sentiment. But if a spin-off is truly imminent, the stakes are far higher than just boosting imagination.

Dream-Making Requires Substantial Investment

Kuaishou's intention to spin off Kling can be seen as a phased validation of Kling's independence. A deeper reason may be that Kuaishou has recognized the nature of the AI video battle—a war centered around infrastructure.

New infrastructure often brings fresh efficiency and possibilities, unlocking capabilities that were previously unattainable and defining a new era.

AI video is likely no exception.

In traditional filmmaking, producing a commercial advertisement involves a lengthy process spanning dozens of roles. Just the initial shooting setup requires photographers, focus pullers, lighting technicians, crane operators, sound engineers, and more, all working in tandem like cogs in a machine.

For the film industry, this industrial pipeline represents decades of evolutionary refinement, ensuring professionalism and upholding visual quality. However, for most people, this hierarchical system acts as a heavy lock, preventing ordinary individuals from realizing their creative visions and keeping the majority out.

AI video models serve as the key to this lock. Once they surpass a usability threshold, no matter how the current film industry resists, criticizes, or sets boundaries, they will struggle to prevent the transfer of creative power.

This is why Kling must fight this battle. The winner will have the opportunity to stand at the gateway of the next-generation content industry.

Although Kling is not currently the most prominent player in AI video, from a long-term perspective, current differences are insignificant.

An industry insider told Photon Planet that the generation quality of video models largely depends on their generational tier. Once models reach the same tier, the capability gap will not be insurmountable. Seedance 2.0's sudden rise was due to its early entry into a new generation, giving it a short-term lead. However, in the long run, the gap between video models will eventually narrow.

In other words, whether it's Kling, Seedance, or Happy Horse, they don't need to lead in every generation. As long as they remain in the game, they have a chance to wait for model capabilities to converge and share in the eventual success.

However, surviving until that day requires vast amounts of money.

At this stage, the industry recognizes that AI is not an internet product that can infinitely acquire users, offer free trials, and eventually rely on scale to reduce costs. Every click behind the generate button burns real money.

Take Seedance 2.0 as an example. As an industry benchmark, it attracted a flood of users to ByteDance's Jimeng, leading to repeated "card-drawing" behavior. According to industry insiders, Jimeng's initial pricing was unsustainable because it underestimated the disruptive power of this "card-drawing economy," forcing it to raise prices despite user backlash.

In other words, for consumer-facing AI video models, user demand is unlimited, while computing power supply is limited. The cost and revenue curves will never intersect.

Sora's shutdown, apart from OpenAI's need to focus, was also due to its failure to find a viable business model amid high costs. Media reports suggest that Sora's daily operational costs reached $15 million, amounting to over $5 billion annually.

While Kling, primarily targeting businesses, can avoid some of the cost black holes caused by consumer "card-drawing" behavior, staying in the game requires continuous payment for computing power, training, and inference.

If it were just about spending money, that would be manageable. However, Kuaishou faces a more realistic and weighty proposition: whether it is willing to bet everything on Kling.

How Long Can the "Love Supply" Last?

Kuaishou has already been generous to Kling.

During last quarter's financial report conference call, Kuaishou CFO Jin Bing stated that the group's total capital expenditures for 2026 are expected to reach around RMB 26 billion, an increase of approximately RMB 11 billion from 2025, primarily allocated to Kling and AI computing power. This quarter, Jin Bing reaffirmed that this guidance remains unchanged.

This expenditure already covers Kuaishou's annual net profit, causing a stir in public opinion.

However, for the AI video battle, this heavy bet can only sustain short-term catch-up, not long-term attrition.

The reason is that infrastructure competition is inherently a winner-takes-all battle—survival or extinction.

In other words, few models will remain in the game long-term. Ultimately, only a handful of video models will become shared infrastructure. Other players, even if they shine briefly, will gradually fall behind due to constraints in computing power, costs, scenarios, ecosystems, and capital.

Facing this increasingly brutal attrition war, if Kuaishou firmly holds onto Kling, it must prepare to pay an even heavier price—not just burning money but potentially shouldering the platform's fate.

The intention behind spinning off Kling is essentially Kuaishou's way of loosening its grip, relieving Kling of cost pressures on itself and allowing Kling to use a more independent and flexible financing narrative to shoulder AI video's long-term investments. It also relieves Kuaishou of the psychological burden of dragging down a "gifted" Kling.

However, past internet giant spin-offs have primarily involved mature businesses like payments, logistics, music, and film—clearly defined segments with relatively stable financial models that can continue operating independently.

Kling is different. Currently, while there are many video generation models, Kling's value stems not only from its first-mover advantage and pure technical and model prowess but also from the continuous support of Kuaishou's content, advertising, transaction, and recommendation ecosystems.

This capability does not belong to an isolated model company.

Therefore, rather than debating "to spin off or not," a more worthwhile discussion is how to partially separate organizationally while maintaining deep coupling in business operations.

After all, for Kling, insufficient separation will lack independent narrative power, making its valuation less attractive. Conversely, excessive separation may lead to pure model competition, pitting it against other players in price, computing power, and delivery wars—likely a grueling battle.

Ultimately, while Kuaishou may have few new stories left to tell, Kling's story is just beginning. Kuaishou's role may be to continue nurturing its growth while learning to let go at the right time.

If a spin-off materializes, Kling could become a case study for how large corporations approach AI. Given the current AI market, where spun-off AI products and business lines from big tech companies can achieve valuations comparable to leading AI startups, the long-term implications of this model for the entire AI sector warrant closer observation.

WeChat Official Account | Photon Planet