Kuaishou, the Chinese short-video platform, has spun off its AI video unit Kling into a separate company and closed an initial $2 billion funding round. The round values Kling at $18 billion after investment. At least 38 investors participated, including Tencent, CITIC Securities, Alibaba Cloud, and Baidu. The round can still grow to $3 billion within 60 days, at which point Kuaishou's stake would fall to 68%. A Hong Kong IPO is being targeted for 2027.
The valuation was trimmed from an original $20 billion ask, which reflects real market pressure even for top-tier Chinese AI fundraises. That $2 billion markdown is not a failure. It is evidence that institutional investors are applying actual discipline to AI pricing, not just writing blank checks based on excitement.
The revenue story is what makes Kling different from most AI fundraising hype. Its annualized revenue run rate grew from roughly $300 million in January 2026 to $500 million by March, with first-quarter revenue up more than 300% year-on-year. About 75% of that revenue comes from overseas markets, primarily from advertising agencies, film production teams, and professional content creators. The most expensive subscription plan is the best-selling one, which tells you the customers are professionals who depend on the tool to earn money, not hobbyists.
Kling launched in June 2024 and has since grown to over 60 million global users and more than 30,000 enterprise clients and developers. Its latest version, Kling 3.0, generates native 4K video, handles multilingual lip-sync from a single text prompt, and lets creators control camera movement and character consistency across multiple shots in a single sequence. Independent reviewers rank it among the top two or three tools in the world for professional video work, alongside Google's Veo 3.1 and ByteDance's Seedance.
The spinoff structure itself is worth noting. Kuaishou's main business is a short-video and e-commerce platform. Kling is an AI production tool sold to businesses and creators. Investors value those two businesses very differently, so separating them lets Kling attract capital and a public valuation on its own terms. Several other Chinese AI companies, including MiniMax and Zhipu AI, have already listed on the Hong Kong Stock Exchange in 2026, with MiniMax doubling on its first day of trading. Hong Kong raised more IPO capital than any other exchange in the first quarter of 2026, pulling in $14 billion across 40 listings.
For Kling, the path to a 2027 IPO is not guaranteed. Its competitor ByteDance has Seedance, which some reviewers rate above Kling on raw cinematic quality. Alibaba has its own AI video tool built by former Kling engineers. And running a video generation business is expensive: Kuaishou plans to spend roughly $3.8 billion on computing infrastructure in 2026, with $1.6 billion of that dedicated to Kling and related AI work. OpenAI shut down its own AI video tool, Sora, after finding it cost roughly $1 million per day to run while generating almost no revenue. Kling has avoided that trap so far, but the cost pressure does not go away at scale.
For non-technical business operators, the practical point is simpler than the funding story. AI-generated video has moved from novelty to production-ready tool. Traditional video production costs around $4,500 per minute; AI-assisted production brings that down to about $400 per minute. The average time to produce a 60-second marketing video has dropped from 13 days to 27 minutes. If your business produces any video content, whether for product pages, internal training, customer communications, or advertising, the cost and time case for running small AI video pilots is already strong. The tools to do this include Kling, but also Google Veo, Runway, and several others. You do not need to pick a winner. You need to start testing.