China's AI Startup Landscape: Who Holds the Greatest Advantage in the Large Model Race?

Since the launch of ChatGPT, China has experienced an AI surge, prompting a wave of entrepreneurs, former senior executives, and academic researchers to launch AI startups focused on large‑model development.

The "entrepreneur" camp can be divided into three groups: the "big‑boss team" (including Wang Huiwen, Li Kaifu, Wang Xiaochuan), the "former tech‑executive team" (e.g., former Alibaba VP Jia Yangqing, former JD AI chief Zhou Bowen), and the "academic team" (e.g., Fudan’s Qiu Xipeng, Tsinghua’s Tang Jie).

These groups pursue two main directions: building general‑purpose large language models to create a Chinese counterpart of OpenAI, or developing vertical, domain‑specific models that integrate tightly with particular applications.

Funding varies widely—Wang Huiwen has invested $50 million of his own money and secured $230 million from top VCs, while Li Kaifu’s Innovation Works backs Project AI 2.0. Talent acquisition remains a bottleneck, with many teams struggling to attract top AI researchers.

Major tech giants are also accelerating their AI efforts: Baidu’s "Wenxin Yiyan", Alibaba’s "MoDa", Tencent’s "Tencent Zhiying", and 360’s ChatGPT‑like demo, each targeting either B‑to‑B cloud services or consumer‑facing products.

Investors note that only a handful of teams can realistically deliver viable large models, emphasizing the importance of vertical applications and ecosystem integration, while the shift toward conversational AI reshapes search advertising and broader market strategies.