By Wang Chunhao

Source: Jiedian Finance

Looking back at the end of 2025, China's artificial intelligence sector is undergoing a paradigm shift.

If 2023 was the wild outbreak of the 'Hundred Models War' and 2024 was the differentiation period of technological routes, then 2025 has undoubtedly entered the most brutal and crucial 'capital realization period.'

With the disillusionment of generative AI technology and the urgent need for commercialization, the once-noisy market quickly cooled and solidified. Three 'unicorns'—Zhipu AI, MiniMax, and Yuezhi'anmian—broke through, forming the first echelon in China's large model sector and collectively launching a final sprint to the Hong Kong Stock Exchange.

Jiedian Finance believes that this is not just a race among three companies but the ultimate showdown between two business models, two technological beliefs, and the capital wills of two internet giants.

MiniMax is becoming the strongest dark horse for its IPO with its strong monetization capabilities in overseas markets and unique product matrix, aiming straight for January 2026. Zhipu AI, as the representative of the 'Tsinghua faction' national team, although abandoned the Science and Technology Innovation Board, quickly shifted to the Hong Kong Stock Exchange, attempting to establish an valuation anchor with an infrastructure stance. Yuezhi'anmian, despite winning user mindshare in the long-text field with Kimi, struggles to balance between computing costs and commercialization, with a relatively delayed IPO timeline.

Meanwhile, the financing window in the primary market is closing. Investors are no longer paying for mere 'parameter scale' but instead demanding a clear 'unit economic model.' An IPO is not just a financing channel but also a 'get-out-of-jail-free card' for these unicorns to survive in the next round of reshuffling.

01 The Battle Among the Top Three: How Are They Performing?

Among the three contenders, MiniMax is the most unique and the most likely to cross the finish line first.

This Shanghai-based company was founded by Yan Junjie, the former vice president of SenseTime. Unlike its competitors with a strong academic atmosphere, MiniMax has had a strong commercial monetization bloodline since its inception.

MiniMax's shareholder list is star-studded, including Alibaba, Tencent, and Sequoia China. Sources revealed that the company plans to list on the Hong Kong Stock Exchange as early as January 2026, aiming to raise hundreds of millions of US dollars. As the currently largest revenue-generating and most clearly defined business model, MiniMax is seen by investment banks as the easiest 'first stock' for the secondary market to understand and accept.

MiniMax's core asset is not some underlying model with hundreds of billions of parameters but a consumer-facing application named Talkie. This app avoids the 'productivity tool' red ocean like ChatGPT and enters the 'AI companion' blue ocean market.

According to Bloomberg, MiniMax's expected revenue in 2024 is as high as $70 million, with the vast majority coming from Talkie. This figure stands out in China's AI startup circle, which generally lacks self-sustaining capabilities.

More critically, 70% of MiniMax's revenue comes from overseas markets, especially the United States. Talkie is extremely popular among American teenagers, consistently ranking high in app store charts. Its user stickiness and willingness to pay far exceed those of similar domestic products. By introducing gamified payment mechanisms like 'gacha,' MiniMax successfully transforms AI computing power into high-margin virtual goods rather than cheap API calls.

If Talkie solves the 'survival' problem, then Hailuo Video embodies MiniMax's development ambitions. In the second half of 2024, MiniMax released the video generation model Hailuo AI, directly competing with OpenAI's Sora and Kuaishou's Kling.

Jiedian Finance analyzes that video generation is considered the next Explosion point (translated as 'breakout point') after text and is also the field with the highest computing power consumption and technological barriers. Hailuo AI quickly topped global video generation product charts after its release, proving that MiniMax not only understands products but also possesses the technological prowess to compete with giants.

If MiniMax is a nimble 'ranger,' Zhipu AI is a steady 'heavy armor battalion.' With its pure lineage from the Knowledge Engineering Laboratory of the Computer Science Department at Tsinghua University, it occupies an irreplaceable ecological niche in China's large model landscape.

Zhipu AI's strategic core is to build an underlying ecosystem similar to the 'Wintel' alliance. By open-sourcing small and medium parameter models, Zhipu quickly captured developer mindshare and then guided enterprise users to migrate to more powerful closed-source commercial models.

According to Tech in Asia and related reports, Zhipu AI's total revenue in 2024 is expected to reach $42 million, with annual recurring revenue from AI development tools exceeding $14 million. This means Zhipu has initially established a SaaS model.

Zhipu's greatest moat lies in its B2G and B2B businesses. Against the backdrop of data security becoming a national strategy, Zhipu has secured numerous privatization deployment orders from governments and state-owned enterprises with its 'autonomous and controllable' label, giving its revenue extremely high certainty and resilience to economic cycles.

Zhipu AI was once the most highly anticipated AI company for the Science and Technology Innovation Board. However, in mid-2025, the company's strategy took a sharp turn, deciding to redirect to Hong Kong. Jiedian Finance believes this decision reflects the management's pragmatism. The Science and Technology Innovation Board's requirements for profitability and research and development capitalization are too rigid for Zhipu, which still requires massive investment. The flexibility of the Hong Kong market and its status as an international capital gateway better align with Zhipu's globalization strategy of 'based in China, radiating the Belt and Road.'

Yuezhi'anmian, founded by the post-90s genius Yang Zhilin, is the youngest and most Silicon Valley-esque among the three. It carries China's tech circle's ultimate fantasy for AGI (Artificial General Intelligence).

Yuezhi'anmian rose to fame with the Kimi smart assistant. Kimi was the first to support ultra-long context input of 200,000 to 2 million words, precisely hitting the pain points of knowledge workers processing financial reports, legal documents, and academic papers.

As of the third quarter of 2025, Kimi's monthly active users are approximately 9 million. Although lagging behind ByteDance's Doubao in absolute numbers, it performs exceptionally well in core metrics like user duration and retention. Its user demographic is mostly highly educated groups with significant potential commercial value.

However, long-text processing incurs high inference costs. Processing 2 million words of context consumes massive video memory and computing power. Under the current environment where C-end users are accustomed to free usage, Kimi faces the dilemma of 'the more it's used, the more it loses.'

Yuezhi'anmian has attempted a 'tipping' model and launched enterprise-grade APIs. However, compared to MiniMax's gamified revenue and Zhipu's B-end contracts, Kimi's revenue-generating capabilities remain weak. This directly led to the delay of its IPO timeline—planned for the second half of 2026, apparently to allow more time to optimize the unit economic model.

02 Alibaba and Tencent's 'Double Down'

Behind this IPO race stand two colossal figures—Alibaba and Tencent.

A deep analysis of the three companies' equity structures reveals that Alibaba and Tencent have invested in all three almost simultaneously.

Unlike the clearly divided camps of US tech giants (Microsoft bound to OpenAI, Amazon bound to Anthropic), China's Alibaba and Tencent have adopted a 'shotgun' approach to cover the entire sector.

Under the leadership of Joe Tsai and Wu Yongming, Alibaba has established AI as its core strategy. Through investments, Alibaba not only gains financial returns but, more importantly, secures future computing power demand for its cloud business. Alibaba's lead investment in Yuezhi'anmian's latest $1 billion financing round is seen as a typical case of this strategy.

Tencent's strategy focuses more on connecting at the application layer. As the social media giant, Tencent does not want to miss out on the next WeChat. Investing in MiniMax and Yuezhi'anmian serves as Tencent's double insurance beyond its own Hunyuan large model.

Jiedian Finance believes that large models could disrupt existing internet entry points, including search, e-commerce, and social media. Since it's impossible to determine the final winner, the safest approach is to buy the entire sector. This means that regardless of whether MiniMax, Zhipu, or Yuezhi'anmian goes public first, Alibaba and Tencent will almost certainly be the winners behind the scenes.

03 The Final Showdown in Commercialization: The Explosion of the Application Layer

In May 2025, ByteDance released the Doubao large model, pricing it by the 'cent' and instantly igniting the market. Zhipu quickly lowered the price of GLM-4.5 Lite to $0.11 per million tokens.

The giants aim to clear the field. For independent startups without cloud provider backgrounds, this is an asymmetric war. Giants can sell models at a loss and profit from cloud services; startups lose money with every model sale.

This forces MiniMax and Yuezhi'anmian to move up the stack and create super apps. At the application layer, users pay through advertising or subscriptions and are insensitive to underlying token costs.

In the second half of 2025, the market landscape gradually became clear.

At the underlying level, Alibaba (Tongyi), Tencent (Hunyuan), Baidu (Wenxin), ByteDance (Doubao), and Zhipu (GLM) have divided the general-purpose large model market.

At the application layer, MiniMax (Talkie) and Yuezhi'anmian aim to become the 'TikTok' and 'WeChat' of the AI era. Data shows that Talkie's user duration already surpasses many traditional social software, proving that AI Native applications have the potential to disrupt old orders.

On the other hand, facing an overly competitive domestic market, overseas expansion is not a choice but a necessity.

The success of MiniMax's Talkie proves Chinese teams' dimensional superiority in ToC products. With its impressive financial data and clear business model, MiniMax is highly likely to ring the bell first in January 2026, becoming China's first large model company to go public. This will also significantly boost industry confidence and set a high valuation benchmark for subsequent IPOs by Zhipu and Yuezhi'anmian.

However, Jiedian Finance believes that going public is just the beginning. Under the spotlight of public markets, every dollar of R&D investment must translate into revenue. For these young entrepreneurs, the real test has just begun. They must not only face technological iterations but also the cold scrutiny of Wall Street and Central investors.

China's AI golden age is accelerating with the support of capital.

*The featured image is generated by AI.