OpenAI Competitors Gauge Interest in Chinese AI Through Dual IPOs

China’s Leading AI Startups Make Their Hong Kong Market Debut

Photographer: Gabby Jones/Bloomberg

Two of China’s most prominent generative AI firms are preparing to launch on the Hong Kong stock exchange this week, offering a glimpse into investor confidence in the country’s rapidly evolving AI sector. Zhipu and MiniMax, both early competitors to global leaders like OpenAI, have taken notably different paths to reach the public markets.

Their listings—Zhipu on Thursday and MiniMax on Friday—will serve as a key indicator of whether investors believe China’s emerging AI industry can stand alongside international giants. Both companies exemplify China’s pragmatic approach to AI, operating with fewer resources and smaller teams compared to the likes of OpenAI and Anthropic. However, their backgrounds and business models set them apart.

Top Stories from Bloomberg

Zhipu, established six years ago by professors from China’s top university, primarily serves domestic clients and enjoys strong local backing. In contrast, MiniMax is a younger startup with a more conventional structure, aiming to rival DeepSeek and OpenAI with its consumer-facing chatbots both within China and internationally. Its supporters include Alibaba Group and Abu Dhabi’s sovereign wealth fund.

MiniMax operates much like a Silicon Valley company, earning most of its income from over 210 million users of its AI-powered apps through subscriptions and ads. Zhipu, which is state-supported, focuses on institutional customers and boasts a client base of more than 8,000 organizations.

“These IPOs are part of a broader policy-driven investment cycle as China ramps up efforts to build a homegrown semiconductor and AI ecosystem,” explained Aadil Ebrahim, head of equities at Klay Group in Singapore. “Investor enthusiasm is evident, as seen in the strong performance of recent GPU and AI-related listings in both the A-share and Hong Kong markets.”

In 2024, Zhipu reported revenue of 312.4 million yuan (about $44.7 million), outpacing MiniMax’s $30.5 million. While these numbers are modest compared to OpenAI’s estimated $13 billion and Anthropic’s projected $9 billion in annual revenue, the valuations of Zhipu and MiniMax are also significantly lower, each around $4 billion before going public.

Zhipu’s team, composed of top AI researchers from Tsinghua University, is recognized as a trailblazer in China’s AI landscape.

The company was founded in 2019 under the leadership of renowned academic Tang Jie. Inspired by OpenAI’s release of GPT-3 in 2020, Zhipu set out to develop an open-source model of similar scale. With a team of 18 scientists and a budget of just 4 million yuan for computing resources, they achieved this goal within two years, according to co-founder and CEO Zhang Peng.

“When we looked at our finances, it was clear that if things went wrong, we wouldn’t have enough funds to finish,” Zhang recalled. “We really only had one chance to succeed.”

Zhipu’s leadership sees the Beijing-based company as a reflection of China’s distinctive approach to AI—closely tied to the real economy and major industries. While Zhang admits that private, on-premise AI deployments may lack the appeal of consumer apps, he argues they are exactly what Chinese enterprises need. State-owned companies are unlikely to use open-source or public cloud solutions for building AI from scratch, creating opportunities for Zhipu’s tailored offerings.

MiniMax, headquartered in Shanghai, has its origins in the gaming world. Co-founder Yan Junjie, an avid Dota 2 player, first noticed OpenAI in 2019 when its bots defeated top human competitors. Fascinated, he immersed himself in OpenAI’s research, shifting his focus from computer vision to natural language processing. Yan pitched investors on creating AI agents capable of passing the Turing test—a vision many doubted. One of his earliest backers was Mihoyo, a gaming studio passionate about integrating AI into games, which remains a key client today.

Both Zhipu and MiniMax are aiming for global expansion, prioritizing growth over immediate profits. Zhipu now offers an AI coding assistant similar to Anthropic’s Claude, but at a much lower price point. MiniMax competes with Sora and a range of Chinese companies in the AI video generation space. Both are expanding into Model-as-a-Service, a cloud-based solution that lets users deploy proprietary models within their own applications.

However, as in the United States, there are concerns in China about heavy investment in AI infrastructure without clear paths to profitability.

After enduring a fierce price war known as the “Battle of One Hundred Models,” both companies are poised to raise substantial funds. MiniMax has attracted cornerstone investors for its IPO, including Alibaba, the Abu Dhabi Investment Authority, and South Korea’s Mirae Asset Securities. Zhipu, limited by US trade restrictions, has secured mainly domestic support from firms like CICC and Taikang Insurance.

Retail demand has been extraordinary: MiniMax’s shares were oversubscribed by more than 1,350 times, while Zhipu’s saw over 900 times the available shares requested, according to margin loan data from the Futubull app.

“While everyone enjoys seeing rapid gains and discussing them, the real measure will be how these companies perform six months or a year from now,” said Philip Wen, managing director at Hong Kong-based MPM Capital. “I still believe that China’s tech giants—ByteDance, Alibaba, Tencent, and Kuaishou—will likely remain dominant in the long run, with smaller firms providing specialized AI solutions.”

The wave of AI IPOs follows a series of high-profile listings by Chinese chipmakers, which are central to China’s ambitions for technological independence and AI leadership. Last month, Moore Threads Technology’s shares soared 425% on their first trading day in Shanghai, followed by a 693% surge for MetaX Integrated Circuits Shanghai.

“This surge of IPOs is crucial for the Chinese government, which is clearly committed to giving these companies access to public capital,” said Ebrahim of Klay Group. “At the same time, investors should be aware that most of these newly listed firms are still unprofitable and may remain so as they scale up.”

Reporting by Luz Ding, Zheping Huang, Sangmi Cha, and Dave Sebastian

With contributions from Sangmi Cha and Dave Sebastian.

(This article includes updates on retail investor demand and additional investor commentary.)

Most Popular from Bloomberg Businessweek

©2026 Bloomberg L.P.