Summary of the Key News
A leading domestic large-scale model company, DeepSeek, has reportedly sought assistance from CITIC Securities to prepare for an IPO on the Science and Technology Innovation Board (STAR Market). Instead of following the paths taken by Zhipu and MiniMax, which have already listed their large-scale models on the Hong Kong Stock Exchange, DeepSeek aims to debut on the A-share STAR Market, where no pure general-purpose large-scale models have previously been listed. The ultimate goal is not just to claim the title of the “first large-scale model company listed on the A-share market” but also to gain control over the local pricing power for the entire general-purpose large-scale model industry in China. In the future, the valuation of all domestic large-scale model companies when they list or seek financing will likely be based on DeepSeek’s public trading price.
---
Detailed Explanation
1. Choosing the STAR Market over the established Hong Kong Stock Exchange
There are three key reasons why DeepSeek has opted for the STAR Market:
- Policy Opportunity: In June this year, the Shanghai Stock Exchange revised its listing rules for large-scale model companies, opening the “fifth set of listing criteria” to companies that rely on technology rather than profit. As long as DeepSeek has a sufficient number of users, strong technology, and clear commercialization prospects, it can list without generating substantial profits. This represents a dedicated entry channel for large-scale models, which was not available on the A-share market before.
- Strong Foundation: DeepSeek currently has 130 million monthly active users and is one of the most popular AI applications in China. It has recently raised two rounds of funding, attracting investors such as Tencent, CATL, and JD.com, totaling over 100 billion yuan. This financial strength, along with its technical capabilities and user base, meets the listing requirements. Additionally, the core engineers and early investors are eager to realize their equity gains after the company goes public.
- Lack of Benchmark in the A-share Market: While the Hong Kong Stock Exchange already has Zhipu and MiniMax as listed large-scale model companies, the A-share market lacks such benchmarks. If DeepSeek succeeds in listing, it will become the natural pricing standard for the industry, similar to how the first Sichuan restaurant in a street sets the price for all other restaurants in that area.
2. Becoming the Industry’s Valuation Standard
Once listed, DeepSeek will establish a clear valuation benchmark for the entire industry. Currently, the valuations of domestic large-scale model companies are determined through private negotiations. However, with DeepSeek’s stock price being determined by the daily trading of tens of millions of investors and institutions, its financial performance will be transparent and publicly available. Investors will use DeepSeek’s data to compare other companies’ models, setting realistic valuations rather than relying on speculative PPT presentations.
3. Impact on Competitors
This impact is already evident:
- Unlisted Companies: Companies like Kimi, which operate in a similar manner to DeepSeek, may see their valuations adjusted downward if DeepSeek’s success sets a lower standard.
- Already Listed Companies: The valuations of Zhipu and MiniMax on the Hong Kong Stock Exchange have diverged significantly. If DeepSeek’s valuation becomes the industry standard, companies with different business models (API-based or application-based) will face different valuations, potentially leading to adjustments in their own valuations.
4. Challenges After Listing
Listing is just the beginning of the real challenges for DeepSeek. The A-share market requires quarterly reports on revenue, losses, and user data, which will be closely monitored by investors. DeepSeek’s performance over the next few quarters will determine whether its previously inflated valuation of 50 billion yuan can be maintained. If its financial results fall short of expectations, its stock price could be significantly adjusted downward.
Risk Warning
This analysis is based on industry logic and does not constitute investment advice. The commercialization of large-scale models is still in its early stages, and the stock prices of related companies are highly volatile. Ordinary investors should not blindly follow market trends.