MiniMax Doubles on Hong Kong Debut, Then Files for Mainland IPO to Chase DeepSeek

MiniMax Group didn't just list in Hong Kong, it exploded onto the board. Shares of the Chinese AI startup doubled on their first day of trading on January 9, raising 4.8 billion Hong Kong dollars, or about $620 million. That performance made it the second major Chinese large language model developer to go public, and it handily outpaced the debut of rival Zhipu AI, which had listed just a day earlier. According to Bloomberg, shares jumped as much as 78 percent in intraday trading.

The offering priced shares between 151 and 165 Hong Kong dollars, with the company selling roughly 25.4 million shares. Investor appetite was voracious. The Standard reported that MiniMax spearheaded a wave of six Hong Kong listings worth a combined 16.7 billion Hong Kong dollars, capping a year in which the exchange raised $36.5 billion from 114 new listings, its strongest performance since 2021. For a startup founded in 2021 by former SenseTime executive Yan Junjie, the reception signals that global capital markets see Chinese AI as a bet worth making right now.

Behind the IPO pop are numbers that tell a story of breakneck expansion. Bloomberg reports that MiniMax's annualized revenue more than doubled over the two months leading up to the listing. The South China Morning Post adds a critical detail: the company's global enterprise and developer client base surged fivefold in six months, reaching one million clients. Its total global user base now sits at 300 million.

Roughly 70 percent of MiniMax's revenue comes from overseas markets, a striking figure for a Chinese AI firm. That international footprint gives the company a narrative that distinguishes it from peers more dependent on the domestic market. The company plans to channel IPO proceeds directly into research and development, according to a person familiar with the matter cited by Yahoo Finance. The goal is straightforward: build the next flagship model and keep the revenue curve pointed up while competing with both global heavyweights and fast-moving local rivals.

The Hong Kong listing was only act one. Bloomberg reports that MiniMax has already begun preparations for a domestic initial public offering on a mainland Chinese exchange, according to a regulatory filing. The move targets a different pool of capital and a different competitive dynamic. While the Hong Kong debut positioned MiniMax against Zhipu AI and other members of China's so-called six little dragons, a mainland listing puts it in direct competition for investor attention with DeepSeek, the Hangzhou-based lab that has captured global headlines for its efficient, high-performance models.

Ainvest notes that MiniMax's projected valuation reached up to $6.5 billion at the time of the Hong Kong pricing. A successful mainland IPO would likely push that figure higher and give the company a dual-listed structure that few Chinese AI startups have attempted. The strategy reflects a broader reality: China's AI race is now a capital markets race, and companies that can tap both international and domestic investors will have a structural advantage in funding the compute and talent wars ahead.

MiniMax is one of a cohort that Chinese tech watchers call the six little dragons: Moonshot AI, Zhipu AI, Baichuan, StepFun, 01.AI, and MiniMax itself. Each is racing to build foundation models that can compete with OpenAI, Anthropic, and Google, while also fending off DeepSeek's aggressive open-source strategy. Zhipu AI listed first, but MiniMax's stronger debut performance reset expectations for what these companies can command in public markets.

The consumer-facing angle matters here. Vincentxiang.substack points to MiniMax's companion chatbot products as a key differentiator. While some rivals focus on enterprise API businesses, MiniMax built a viral consumer product that gave it brand recognition and a user base that enterprise sales alone can't generate. That dual consumer-enterprise approach mirrors what OpenAI attempted with ChatGPT, and it appears to be resonating with retail and institutional investors alike. The question now is whether MiniMax can sustain that momentum when DeepSeek's models are widely seen as matching or exceeding the performance of far more expensive systems.

MiniMax's listing didn't happen in isolation. The Standard reports that six Hong Kong IPOs launched on the same day, with AI and chipmaker companies driving the city's equity markets comeback. After a sluggish 2024 that raised roughly $11.3 billion, Hong Kong pulled in more than triple that amount in 2025. The exchange is positioning itself as the listing venue of choice for Chinese tech companies that face geopolitical friction with U.S. exchanges.

For MiniMax, Hong Kong offers access to international capital without the delisting risks that Chinese firms face on the Nasdaq. The company's 70 percent overseas revenue share makes that international investor base particularly valuable. But the planned mainland IPO suggests that MiniMax doesn't want to choose between global capital and domestic depth. It wants both. That dual-exchange strategy could become a template for other Chinese AI firms watching from the sidelines, especially if MiniMax's mainland listing matches the enthusiasm of its Hong Kong debut.

For all the revenue growth and user metrics, MiniMax is not profitable. The South China Morning Post reports that the company recorded large total losses even as its annualized revenue doubled. That's the reality of the foundation model business: training runs cost tens of millions of dollars, inference at scale is expensive, and pricing pressure from open-source competitors like DeepSeek squeezes margins. The IPO cash buys runway, not a path to profitability.

Decodingdiscontinuity frames the listing as a window into the economics of intelligence itself. The market is effectively pricing what it costs to produce and sell access to advanced reasoning at scale. MiniMax's valuation, at up to $6.5 billion, reflects a bet that the company can eventually convert its 300 million users and one million enterprise clients into a sustainable business. Whether that bet pays off depends on the next model launch, which Bloomberg reports is imminent, and on whether MiniMax can keep its revenue doubling while competitors race to undercut it on price.