Summary of Key Points
Kuaishou’s video AI company, KelingAI, has secured nearly $3 billion in financing, setting a new record for the largest video AI model investment globally. The rare simultaneous investment from BAT (Baidu, Alibaba, and Tencent) has been interpreted as an “anti-Baidu alliance,” but in reality, each company is pursuing its own financial objectives (with a combined stake of only 2.51%). By spinning off KelingAI for independent financing, Kuaishou aims to leverage the high valuation of AI to boost its stock market value (KelingAI’s valuation accounts for over 70% of Kuaishou’s total market value) and counter the slowdown in growth of its main business. Although KelingAI has shown rapid growth (300% revenue increase in Q1 2026) and industry recognition (winning awards at Cannes), it faces challenges such as weak technical barriers, competition from ByteDance in overseas markets, and a significant gap in computing power. The financing also includes a conditional clause requiring KelingAI to go public by 2031; otherwise, Kuaishou must repurchase the shares.
I. BAT’s Joint Investment is Not an “Anti-Baidu Alliance”
The market suggests that BAT is working together to “contain” ByteDance, but upon closer inspection, it’s clear the three companies have different motivations:
- Alibaba: Seeking profits from its cloud services (KelingAI is a major consumer of computing power, and using Alibaba Cloud can generate stable revenue). Additionally, Alibaba wants to use KelingAI to reach overseas customers; it lacks short-video traffic and hopes KelingAI’s channels can help sell its models and cloud services.
- Tencent: Its own AI video model (Zhuying) is lagging behind in terms of image quality and dynamic control. Investing in KelingAI allows Tencent to quickly acquire mature video generation capabilities, reducing costs for its video platforms, advertising, and gaming services, while also competing with ByteDance’s presence in the AI video space.
- Baidu: Its Wenxin AI model excels in text processing but struggles with video. Investing in KelingAI is a way to address this weakness (for $341 million, it obtained video API capabilities) and improve its multi-modal ecosystem for AI search and marketing.
The key point is that all three companies hold very low stakes (Tencent 1.12%, Alibaba 1.11%, Baidu 0.28%) and have no board seats or veto power. This investment is purely financial; they only temporarily align on targeting ByteDance, but in other areas like cloud services and AI models, they still compete with each other.
II. Kuaishou’s Spin-off of KelingAI: A “Clever Strategy” to Boost Valuation
Kuaishou is facing difficulties: its main short-video business is growing slowly (only a 3.4% increase in total revenue in Q1 2026), and its market value has dropped by nearly 90% from its peak, with a price-to-sales ratio lower than that of many traditional manufacturing companies (1.2 times). The spin-off of KelingAI is a last resort:
- After the investment, KelingAI’s valuation is $18 billion, accounting for over 70% of Kuaishou’s total market value of approximately $25.4 billion. However, KelingAI’s annual revenue in 2025 was only 1.04 billion yuan, or 0.73% of Kuaishou’s total revenue—meaning a business that accounts for less than 1% of the parent company’s revenue is supporting a significant portion of its market value.
- The reason is simple: the capital market prefers stories of high AI growth rather than traditional models based on traffic monetization. By separating KelingAI and using the valuation logic of AI businesses (e.g., ARR, Annual Recurring Revenue), Kuaishou hopes to regain investor confidence and stabilize its stock price.
However, this strategy carries risks: KelingAI’s valuation is based on high growth expectations; if future performance fails to meet these expectations, the bubble could burst.
III. KelingAI’s Growth Highlights and Challenges
KelingAI is not without strength, but overcoming ByteDance and overseas giants is no easy task:
- Highlights:
- Rapid growth: Revenue in Q1 2026 exceeded 650 million yuan, a 300% increase year-over-year; ARR increased from 100 million to nearly $500 million ( quadrupling in one year), making it one of the few AI applications with annual recurring revenue over $300 million.
- Industry recognition: KelingAI’s videos won silver and bronze awards at the 2026 Cannes Creative Festival, marking the first time an AI video tool has received an award there, indicating professional creators’ acceptance.
- Challenges:
- Weak technical barriers: Although KelingAI ranks third in third-party rankings (behind ByteDance’s Seedance 2.0 and Alibaba’s HappyHorse), it faces significant talent loss. The replacement of the initial model developer, Zhang Di, by someone from Alibaba resulted in a more advanced model, showing that technology relies on expertise rather than just funding.
- Competition from ByteDance in overseas markets: 70% of KelingAI’s revenue comes from overseas, where ByteDance’s Seedance 2.0 has over 1 billion monthly active users and lower customer acquisition costs. Additionally, with KelingAI’s average monthly price at $127.99, there is little room for price increases, and users may be attracted by competitors.
- Computing power gap: ByteDance plans to invest $200 billion in AI computing power in 2026, while Kuaishou only invested 26 billion—this disparity affects model iteration speed and makes it difficult to compete with ByteDance.
IV. The “Safety Net” Behind the Financing and Long-Term Challenges
The financing includes a conditional clause: if KelingAI does not go public by October 2031, investors can request the company to repurchase their shares at the original amount plus an annual interest rate of 8%. This indicates that:
- Investors are cautious, providing funding but seeking protection against failure (they get a return on their investment with potential for profit or a refund if it fails).
- Both Kuaishou and investors realize that KelingAI will not go public immediately and need five years to achieve growth and validate its valuation.
For KelingAI, the independent financing gives it a chance to continue, but it must face challenges such as ByteDance’s global expansion, competition from overseas giants (like Google), and the pressure of continuous technological improvement. It also needs to balance research and development with profitability, as its valuation is based on promises that must be fulfilled with real growth.
Conclusion
The so-called “anti-Baidu alliance” is just a market rumor; BAT is each pursuing its own interests. Kuaishou’s valuation reconstruction is a practical need, but KelingAI’s future lies in improving its models, retaining users, and establishing a foothold in overseas markets. The financing is just the beginning of a more challenging journey.