The Largest AI Unicorns from China: Who Could Be the Best ‘Chinese ChatGPT’ Besides DeepSeek?

Interest in artificial intelligence models, sparked by the launch of the American ChatGPT in late 2022, continues to grow worldwide. Besides DeepSeek, which shook the world last week, China has produced many promising AI ‘tigers’, each with an estimated value exceeding one billion dollars, and four of them surpassing 2.7 billion dollars in market value. TechNode has highlighted six companies expected to excel in the AI field: Zhipu AI, MiniMax, Baichuan AI, Moonshot, StepFun, and 01.AI unicorns that could replicate the success of American OpenAI, the creator of ChatGPT, on Chinese soil.

What they have in common is that they have attracted the interest of users and investors in a very short time since their establishment, and their teams consist not only of experts from the real sector but also numerous PhDs. Furthermore, in addition to competing to develop a better model, each of them is finding ways to achieve better solutions with minimal resources (money), fiercely competing against more expensive models from the West. Just like DeepSeek, which Forbes also estimates to be worth over one billion dollars, reportedly developed its large language model for just six million dollars and instantly became the biggest competitor to ChatGPT.

Zhipu AI

Founded in 2019, Zhipu AI is one of the first Chinese companies to begin exploring large language models and is now regarded as a major competitor in the field of artificial intelligence. The company is led by numerous leaders from the Chinese academic and business community, such as former Alibaba Cloud Vice President Chen Xuesong. In October 2024, the company released GLM-4.0, a large language model capable of replicating human interactions and offering users the ability to customize tone, emotions, or dialects. In 2023, it attracted investments of 350 million dollars from the Zhongguancun Innovation Fund, Alibaba, Tencent, Xiaomi, and other major companies. Last year, the company secured investments worth as much as 420 million dollars.

MiniMax

Before founding MiniMax in 2021, Yan Junjie, the company’s founder, spent over a decade in artificial intelligence research and development. In October 2022, MiniMax launched its first application, Glow, which attracted five million users in just four months. The application was later renamed Xingye for the Chinese market and Talkie for global markets. In March 2024, Alibaba invested 600 million dollars in MiniMax.

Baichuan AI

In less than a hundred days after its establishment in March 2023, Baichuan Intelligence released two open-source models, Baichuan-7B and Baichuan-13B. Its team consists of AI experts from renowned tech companies such as Sogou, Baidu, Huawei, Microsoft, ByteDance, and Tencent. By October 2023, just over half a year after its establishment, technology giants Alibaba, Tencent, Xiaomi, and several large funds had invested in the company. Last year, Baichuan Intelligence attracted an investment of another five billion yuan.

Moonshot

Another company founded in March 2023 – Moonshot – boasts a team from tech companies such as Google, Meta, and Amazon. Among the available models on the market, Moonshot’s large model Kimi Chat stands out by supporting the generation of a much larger number of characters – approximately 200,000 Chinese characters. Thanks to this, Moonshot is now valued at around 3 billion dollars, although its success has recently been overshadowed by legal disputes with investors.

StepFun

Many people first heard of the company only in March last year, as it quickly became a leading player in the field of large language models shortly after its establishment. The strong multimodal technology they developed attracted wide attention from developers. In the past ten months, the company has launched a total of 11 models, including language models, multimodal understanding models, video generation models, and speech models. Co-founder and CEO Jiang Daxin was a long-time global partner at Microsoft, vice president, and former chief scientist before venturing into entrepreneurship.

01.AI

Kai-Fu Lee, the founder of 01.AI, is also a former top manager at big tech companies such as Apple, Silicon Graphics, Microsoft, and Google. In the first nine months of last year, 01.AI launched two models – Yi-Lightning and Yi-Large – both of which were among the best in China. The key advantage of 01.AI lies in the training cost of these models. Training Yi-Lightning required only 2,000 graphics cards and lasted just a month and a half, with costs amounting to about 2 percent of those for Musk’s xAI. Last year, the company raised several hundred million dollars in investments, and it has held unicorn status, meaning a value exceeding one billion dollars, since a previous investment from Alibaba Cloud.