Kuaishou wants investors to value Kling like a finished AI business, not a flashy demo. The revenue numbers give it a real argument, but the $20 billion price still asks you to believe the growth curve survives both OpenAI and Washington.
Kling is no longer just Kuaishou's answer to Sora. According to The Wall Street Journal, the Beijing short-video company is preparing to spin out its generative AI unit, raise about $2 billion from outside investors and list it in Hong Kong as soon as 2027. Tencent, already a major Kuaishou shareholder, is among the names in early talks. The number attached to the whole exercise is the one you can't ignore: as much as $20 billion.
That would sound breathless if Kling were only a model-release story. It isn't. The Journal reported that Kling's annual recurring revenue climbed from about $150 million in December 2025 to roughly $500 million by May 2026. Kuaishou also said in a May filing that it was assessing a restructuring of Kling that could involve external funding, while making the dry but important point that the proposal was still preliminary and might not proceed.
Keep that caveat in. It is the difference between a deal and a plan.
The pitch to investors is plain enough. Kling has found paying users faster than most AI video companies have found a business model. It makes images and video for advertising, social content and film work, and it has built an international user base in markets including the US, Europe and Japan. If you're looking at AI video as a category, you can get lost in model benchmarks and cinematic demo clips. Kling's better argument is less glamorous: people are paying for the product now.
Runway is the obvious comparison, but it shouldn't be stretched past usefulness. The New York AI video company raised $308 million in April 2025 at a valuation of more than $3 billion, according to public reports at the time, and it remains one of the best-known Western names in generative video. Kling is asking investors to price it far above that. The case for doing so rests on revenue, not taste. A model that looks better in a launch video is interesting. A model running at a $500 million revenue pace is a different conversation.
Pricing is part of that conversation too. Kling has pushed hard on availability and cost, while OpenAI, Google and Runway fight for attention at the high end of video generation. That is a very Chinese internet-company move: build distribution, lower the barrier, and let usage do some of the work the brand cannot. You don't have to admire the strategy to understand it. If the revenue figures reported by the Journal are right, the strategy is already working.
But a $20 billion valuation still needs discipline. At $500 million of annual recurring revenue, Kling would be valued at about 40 times ARR. If investors are underwriting the company on 2027 revenue, the multiple comes down, but only if growth keeps moving at its current pace. That is the bet. Not that AI video will be useful. It already is. The bet is that Kling can keep converting curiosity into paid usage while Sora, Google's Veo, Runway and ByteDance keep shipping better tools into the same market.
There is also the China risk, and investors shouldn't pretend it is background noise. Kuaishou is headquartered in Beijing, and Kling has grown during a period when US export controls have made advanced AI hardware access a moving target. Tom's Hardware reported in January 2026 that the US Commerce Department had created a case-by-case path for exports of Nvidia H200 and AMD MI325X chips to China and Macau, but under strict limits on volumes, end users and testing. That is not a clean supply environment. It is a permission structure.
Frankly, that is where the valuation gets harder. Kling is not on a US entity list, and Kuaishou is not in the same position as Chinese firms already boxed in by direct sanctions. Still, any Western fund looking at this round has to price in the possibility that the rules change before the Hong Kong listing window opens. A business can be growing beautifully and still trade at a discount because the exit path depends on regulators who do not care about your spreadsheet.
The spinout itself makes sense. Chinese tech groups have been carving out AI units because a standalone company is easier to value, easier to fund and easier to take public than a product line buried inside a larger platform. Kuaishou's main app is a short-video and commerce business. Kling is an AI infrastructure and creative-tool business. Investors will pay different multiples for those, and management knows it.
So the question isn't whether Kling is real. The revenue has already made that argument for it. The question is whether Kuaishou can turn a fast-growing AI product into a public company before the market cools, competitors catch up on pricing, or export politics make the story too complicated for global capital. That is why this deal is interesting. It is not obvious, and at $20 billion it cannot afford to be half right.
Also read: HSBC puts a hard number on AI and the banking industry will have to reckon with it • The US government just gave DeepSeek a quiet pass while blacklisting more than 100 other Chinese firms • How to Automate Business Operations with AI Agents Without Hiring an Engineer
