Summary of Key Points
Zhipu (02513.HK) and MiniMax (00100.HK), which were listed on the Hong Kong Stock Exchange in January this year and are known as the "two giants of large-scale models," have initiated the process to list on China's STAR Market (A-share market) in less than five months. Zhipu plans to raise 15 billion yuan for large-scale model research and development, while MiniMax has already begun IPO counseling. This approach of "listing in Hong Kong first and then on the A-share market" not only takes advantage of the international openness of the Hong Kong Stock Exchange to obtain initial capital but also leverages local resources and policy benefits through the A-share market, marking a significant milestone for these large-scale model companies in terms of capital and strategic advancement.
Detailed Analysis
1. Progress of Both Companies Returning to the A-share Market: The STAR Market as a Common Goal
- Zhipu: Has entered the substantive phase of A-share counseling. It began A-share counseling in April 2025, changed its counseling institutions to Guotai Haitong + CICC in February this year, and on June 1, the board of directors officially resolved to apply for listing on the STAR Market. The company plans to issue new shares accounting for 2%-8% of the total share capital (with a maximum oversubscription of 44.58 million shares), raising 15 billion yuan: 12 billion yuan will be used for general large-scale model research and development, 2 billion yuan for building a MaaS (Model as a Service) platform to provide companies with large-scale model tools, and 1 billion yuan for working capital.
- MiniMax: Signed a counseling agreement with CITIC Securities on May 29 and announced its target market as the STAR Market on May 31. However, the specific plan still requires approval from the board of directors and shareholders, as well as consideration of regulatory and market conditions.
The fact that both companies are eager to return to the A-share market within less than five months of their Hong Kong listings highlights their emphasis on the domestic capital market.
2. Why Choose the "Hong Kong First, Then A-share" Path?
In simple terms, it's about finding the most suitable market at each stage:
- List in Hong Kong First: To attract international capital. Large-scale model companies often incur significant losses in their early stages (Zhipu lost 4.7 billion yuan, MiniMax lost $18.7 million). The traditional A-share review process has more restrictions on unprofitable companies, while the Hong Kong Stock Exchange is more accommodating to tech companies, allowing them to secure long-term global capital and establish an "international price benchmark" for their technological value (for example, Zhipu's market value has grown rapidly after listing, and it has been included in the Hang Seng Technology Index).
- Return to the A-share Market: To strengthen local ties. Although Hong Kong is international, domestic computing power, data resources (essential for large-scale model training), and policy benefits (such as government support for AI) can only be fully utilized through an A-share listing. Additionally, the A-share market offers more "patient capital" willing to invest in long-term growth.
Professor Tian Lihui describes this as the "optimal solution of trading time for space": using the breadth of the Hong Kong Stock Exchange to validate global competitiveness and the depth of the A-share market to solidify a firm foundation.
3. Core Motivations for Returning to the A-share Market
- Lack of Funds for Research and Development: Large-scale model development is extremely costly, involving millions or even hundreds of millions in expenses (such as purchasing GPU servers and hiring top engineers). Both companies are growing rapidly but are incurring substantial losses; returning to the A-share market provides an additional funding channel for continued research and development.
- Enhancing Influence: Listing on the A-share market builds trust with domestic customers and partners, expands market share for their models (e.g., Zhipu's GLM series), and increases their presence in the domestic AI industry.
- Strategic Alignment: Returning to the A-share market aligns with national AI strategies, making it easier to obtain policy support and access key local resources such as computing power and data, thus establishing a legitimate position in the industry.
4. Impact on the Market
- A-share Market: Addition of Hard Technology Companies: The A-share market previously lacked companies at the foundational level for large-scale models. The listing of these two giants will direct investment from application-oriented companies (e.g., those using large-scale models for chatbots) towards core technology development.
- Valuation Alignment: Pricing on both markets will be more balanced, with the A-share market placing greater emphasis on technological advantages, commercialization capabilities, and scarcity. In the long run, price fluctuations between Hong Kong and the A-share market will help correct any distortions in valuation.
- Impact on the AI Sector: The listing of these companies will boost the AI sector, as companies providing computing power (GPUs, servers), data, and other necessary resources for large-scale models may see their valuations re-evaluated.
5. Current Company Status: Rapid Growth Amid High Losses
- Rapid Revenue Growth: Zhipu's total revenue in 2025 was 724 million yuan (a 131.9% increase), while MiniMax's revenue was $79 million (a 158.9% increase, with 70% coming from international markets).
- High Losses: Both companies are incurring significant losses, mainly due to heavy investment in research and development (e.g., purchasing servers and hiring talent). This is a common phenomenon for early-stage tech companies, as they need to invest heavily to build technical barriers before achieving profitability.
In summary, the "Hong Kong first, then A-share" approach adopted by Zhipu and MiniMax represents a smart strategy for Chinese large-scale model companies to balance globalization and localization. It allows them to access international capital while leveraging domestic resources. For the market, this move adds valuable hard technology assets, enhancing the vitality of the AI sector. In the future, this path may become the mainstream choice for large-scale model companies.