According to Bloomberg Business, Chinese generative AI startup MiniMax Group Inc. has raised $619 million in its Hong Kong initial public offering. The company, backed by Alibaba and Abu Dhabi’s sovereign wealth fund, sold 29.2 million shares at HK$165 apiece in an upsized offering. It’s set to begin trading soon, following rival Zhipu’s lackluster debut just last Thursday. MiniMax, founded in early 2022, posted an adjusted loss of about $186 million in the first nine months of 2025. Its first-day performance will serve as a critical litmus test for investor enthusiasm in China’s AI software sector.
The Software vs. Hardware Test
Here’s the thing: China‘s stock market has gone absolutely bonkers for AI hardware companies. We’re talking about chipmakers like Moore Threads and Biren Technology surging multifold or 76% on their first day. That’s the power of localization demand and Beijing’s push for semiconductor independence. But software? That’s a different story. Zhipu’s debut was, frankly, a bit of a dud, closing only 13% higher. So MiniMax isn’t just listing a company; it’s testing a whole thesis. Can a pure-play AI software firm, one that’s burning serious cash on R&D and consumer chatbots, generate the same investor frenzy as the companies making the physical chips? I’m skeptical. Hardware feels tangible, a direct bet on geopolitical trends. Software success is fuzzier, dependent on user adoption, monetization, and brutal competition.
Roots in Gaming, Bet on Chatbots
The MiniMax story is actually pretty cool. It has deep roots in gaming, which makes sense. Co-founder Yan Junjie was a Dota 2 player who got hooked on AI after seeing OpenAI’s bots beat human pros. He pivoted his entire career from computer vision to natural language processing. And get this—one of his first investors was Genshin Impact creator Mihoyo, which remains a key client. So you’ve got a company born from a gamer’s obsession, funded by a game studio obsessed with AI. Now, they’re trying to take on DeepSeek and OpenAI with consumer chatbots both in China and abroad. That’s an incredibly ambitious, and expensive, path. That $186 million loss in nine months? That’s the cost of playing in the big leagues. The question is whether that narrative—a gamer’s dream to build a global AI contender—resonates more with investors than Zhipu’s more enterprise-focused pitch.
A Wave of AI Listings
Don’t look at MiniMax in isolation. This is part of a flood. About half of the 11 companies planning to list in Hong Kong this month are in AI. They’re all rushing to raise capital while the window is open, fueled by government support for homegrown champions. It’s a land grab. But as Bloomberg Intelligence analyst Marvin Chen noted, it’s still early in the cycle. It’s really hard to pick winners and losers right now. Basically, we’re about to see a whole cohort of companies hit the public markets, all with similar stories of vast potential and even vaster losses. The performance will start to differentiate them. A strong debut for MiniMax could open the funding taps wider for other software-focused AI firms. A weak one might send them all back to the drawing board, or back to private investors. For enterprises and developers betting on these platforms, it adds another layer of uncertainty. Are you building on a foundation that will be well-funded for the long haul, or one that might struggle to keep the lights on?
What It Means for the Market
So what’s the real impact? For the broader tech market, it’s a signal. A successful listing proves there’s durable, public market appetite for the speculative, cash-intensive model of generative AI software development outside of Nvidia and the chipmakers. It validates the sector. For users, it means the consumer chatbots from MiniMax and its rivals likely aren’t going anywhere soon—they’ll have war chests to keep improving (or at least trying to). But let’s be real: the pressure will now be immense. They’re no longer a plucky startup; they’re a public company with quarterly results to report. The focus will violently shift from user growth at any cost to a path, any path, to profitability. That often leads to more aggressive monetization, pared-back free services, or a sharper pivot to enterprise sales where the money is. The wild, spend-heavy experimentation phase? It might be coming to an end.
