Kling AI, the video generation business unit of Chinese short-video platform Kuaishou, has selected banks for a potential Hong Kong initial public offering that could raise at least $1 billion, according to TechNode reporting on October 6. The Kling AI Hong Kong IPO represents one of the most significant generative AI liquidity events scheduled for the coming year, and would mark the first major public listing of a standalone video generation model provider.
Key takeaways
- Kling AI has selected China International Capital Corp., Goldman Sachs and UBS as underwriters for a Hong Kong IPO targeting at least $1 billion
- The listing is planned for as early as 2027, though the timeline and fundraising amount remain subject to change
- Kling AI raised $2.8 billion in July 2024 at a pre-money valuation of approximately $15 billion, with backing from Alibaba, Tencent and Baidu
- The company currently offers the Kling 2.5 Turbo Pro model at pricing starting from $0.07 per second
- Goldman Sachs and UBS declined to comment on the IPO plans, while Kling AI and CICC did not respond to requests for comment
Banking syndicate and IPO structure
The underwriting syndicate for the Kling AI Hong Kong IPO brings together three major investment banks spanning Chinese and international markets. China International Capital Corp., the Beijing-based investment bank, joins Goldman Sachs and UBS in leading the offering. According to TechNode, discussions remain ongoing and both the fundraising amount and timetable could shift before the listing proceeds.
The $1 billion target represents a conservative floor for the offering rather than a fixed amount. At the company’s last private valuation of approximately $15 billion following its July funding round, a $1 billion raise would represent roughly 6.7% dilution assuming no valuation step-up. However, the actual pricing and size of the offering will depend on market conditions in 2027 and the company’s growth trajectory over the next 12-15 months.
When contacted about the IPO plans, Goldman Sachs and UBS both declined to comment, while Kling AI and China International Capital Corp. did not respond to requests for comment, TechNode reported. This silence is typical during the early stages of IPO preparation, when companies remain under regulatory quiet periods and banks face strict communication restrictions.
Recent funding and valuation context
Kling AI secured $2.8 billion in a July 2024 funding round that established the company’s pre-money valuation at approximately $15 billion. The round drew participation from China’s three largest internet companies: Alibaba, Tencent and Baidu. This investor lineup reflects both the strategic importance of video generation technology and the competitive dynamics among Chinese tech platforms seeking to secure access to cutting-edge generative AI capabilities.
The valuation multiple implied by that funding round positioned Kling AI among the most valuable standalone generative AI companies globally. For comparison, the valuation came during a period when video generation technology was rapidly advancing but had not yet achieved the same commercial penetration as text-based large language models. The backing from Alibaba, Tencent and Baidu suggests these platforms view video generation as a critical capability for their own consumer and enterprise products.
The decision to pursue a public listing roughly 18 months after that funding round indicates confidence in the company’s revenue trajectory and market position. It also suggests that Kuaishou, the parent company, sees strategic value in establishing Kling AI as an independent public entity rather than maintaining it as a wholly-owned subsidiary.
Kling AI’s current product and market position
Kling AI currently offers the Kling 2.5 Turbo Pro model through its commercial API, with pricing starting from $0.07 per second of generated video. This positions the model in the middle of the competitive landscape for video generation services. For reference, Convly’s AI models database shows that ByteDance’s Seedance 2.5 starts at $0.097 per second, while Google’s Veo 3.1 and Alibaba’s Wan 2.5 both begin at $0.05 per second.
The company competes in a market that has seen rapid evolution over the past 18 months, with major players including OpenAI (whose Sora 2 and Sora 2 Pro models were retired in September 2024 when the company shut down its Videos API), Google (Veo series), ByteDance (Seedance), and Alibaba (Wan). The competitive landscape reflects both the technical challenges of video generation and the substantial compute resources required to serve these models at scale.
Video generation models face significantly different economics compared to text-based large language models. Where text models are typically priced per million tokens (with leading models ranging from $0.14 to $10 per million tokens depending on capability tier), video models incur costs that scale with video length, resolution, and generation speed. This creates both opportunities and challenges for companies seeking to build sustainable businesses around video generation technology.
Hong Kong as an IPO venue
The choice of Hong Kong for the Kling AI listing reflects several strategic considerations. Hong Kong has emerged as the preferred venue for Chinese technology companies seeking public listings, particularly in the wake of increased regulatory scrutiny of Chinese companies listing in the United States. The Hong Kong Stock Exchange has actively courted technology companies and has established frameworks specifically designed to accommodate high-growth, pre-profit technology businesses.
For AI companies specifically, Hong Kong offers several advantages: proximity to mainland Chinese investors who understand the domestic market dynamics, a regulatory environment familiar with Chinese corporate structures, and the ability to trade in Hong Kong dollars while maintaining operational flexibility in renminbi. The exchange has also demonstrated willingness to approve IPOs of companies that remain loss-making but show strong revenue growth and clear paths to profitability.
The 2027 timeline would give Kling AI roughly two full years of commercial operations under its belt before going public, assuming the company began material commercial operations in early 2025. This timeline aligns with typical IPO preparation cycles for technology companies, which generally require 12-18 months of preparation after making the decision to pursue a listing.
Implications for the generative AI financing landscape
A successful Kling AI Hong Kong IPO would represent a significant milestone for the generative AI sector, providing the first major public market test of investor appetite for standalone video generation companies. While several large language model providers have discussed IPO plans, and some AI infrastructure companies have gone public, no pure-play video generation company has yet completed a major public listing.
The IPO would also provide important price discovery for the video generation market. Private market valuations for AI companies have shown significant volatility, and public market investors will apply different valuation frameworks than venture capital and strategic investors. Key metrics that public investors are likely to focus on include revenue growth rates, gross margins (particularly important given the compute intensity of video generation), customer acquisition costs, and the company’s competitive position relative to both Chinese and international rivals.
For developers and enterprises evaluating video generation providers, a public listing would bring increased financial transparency. Public companies must disclose detailed financial statements, revenue breakdowns, and risk factors that private companies can keep confidential. This transparency can help customers assess the long-term viability and investment trajectory of their vendor partners. Teams comparing options can use Convly’s AI API cost calculator to model the economics of different video generation services at scale.
Technical and operational considerations
Video generation models require substantially more computational resources than text-based models, both for training and inference. A single minute of generated video can require hundreds or thousands of times more compute than generating an equivalent amount of text. This creates significant infrastructure requirements and capital intensity for companies operating in this space.
The business model implications are substantial. While text model providers can achieve gross margins of 70-80% or higher once they achieve scale and optimization, video generation companies face inherently higher costs of goods sold. The infrastructure required to serve video generation at scale includes not only GPU compute for the generation process itself, but also significant storage and bandwidth costs for delivering the resulting video files to customers.
These economics help explain why video generation pricing remains significantly higher than text generation on a per-second or per-request basis, and why companies like Kling AI require substantial capital raises to build and operate their infrastructure. The $2.8 billion raised in July likely funds not only ongoing model development and commercial expansion, but also the buildout of inference infrastructure capable of serving video generation at commercial scale. Organizations evaluating whether to use API-based services or deploy models internally can assess the tradeoffs using Convly’s self-hosting vs API calculator.
Frequently asked questions
When is the Kling AI Hong Kong IPO expected to occur? The IPO is targeted for as early as 2027, according to TechNode’s reporting. However, the timeline remains subject to change based on market conditions and the company’s preparation progress.
How much is Kling AI trying to raise in its IPO? The company is targeting at least $1 billion in the Hong Kong listing, though the final amount could change as discussions continue and market conditions evolve.
Which banks are underwriting the Kling AI IPO? China International Capital Corp., Goldman Sachs and UBS have been selected as underwriters for the offering, according to TechNode.
What is Kling AI’s current valuation? The company was valued at approximately $15 billion on a pre-money basis in its July 2024 funding round that raised $2.8 billion.
How does Kling AI’s pricing compare to competitors? Kling 2.5 Turbo Pro starts at $0.07 per second, positioning it between ByteDance’s Seedance 2.5 ($0.097 per second) and more affordable options like Google’s Veo 3.1 and Alibaba’s Wan 2.5 (both $0.05 per second).
The bottom line
The Kling AI Hong Kong IPO represents a significant test of public market appetite for generative video technology companies. With a target of at least $1 billion and a tentative 2027 timeline, the listing would provide the first major price discovery for the video generation sector in public markets. The company’s July 2024 valuation of $15 billion and its backing from Alibaba, Tencent and Baidu demonstrate strong private market confidence, but public investors will apply different frameworks focused on revenue growth, margins, and competitive positioning.
For the broader AI industry, the IPO’s progress will offer important signals about investor appetite for capital-intensive generative AI businesses, the sustainability of current private market valuations, and the long-term commercial viability of video generation as a standalone business versus a feature integrated into larger platforms. The selection of leading underwriters and the Hong Kong venue both indicate serious preparation for a major listing, though the 2027 timeline leaves substantial room for market conditions and company performance to influence the final outcome.
Sources: technode.com. Reported October 08, 2026.