2024 Analysis of China’s Large‑Model “Six Tigers”: Slowing Model Gains, Funding Rounds, and Market Strategies

ChatGPT’s breakthrough sparked a wave of generative‑AI entrepreneurship, and in China companies such as Zhipu AI, Baichuan, Zero‑One, Moon of the Dark Side, and Minimax quickly took the lead, with the venture‑backed Step Star receiving a $2 billion round.

The “six tigers” have formed a competitive landscape, and the AI‑application layer is expected to explode this year. In March, Kimi’s 2‑million‑word long‑text model gained massive attention, prompting a fierce user‑acquisition war among model providers.

All six firms secured financing exceeding one hundred million yuan in 2024, yet model capability growth appears to have plateaued, and the community shows fatigue toward hype surrounding new releases such as OpenAI’s upcoming “Strawberry” model, which is positioned as a domain‑specific patch rather than a general‑capability leap.

Domestic leaders have slowed model upgrades, while newcomers like Baichuan’s Yi‑Large and Zero‑One’s Baichuan 4 have begun to match or surpass GPT‑4 in certain tasks. OpenAI’s multimodal GPT‑4o introduced the “Her” feature, sparking a multimodal race that saw Step‑Star’s Step‑1.5V and Zhipu’s GLM‑4‑Plus catch up by late August.

With model improvements stalling, companies are focusing on how to monetize AI capabilities. Moon of the Dark Side’s Kimi pursued a C‑side strategy with long‑text and browser‑plugin features, while MiniMax and Step‑Star launched productivity tools (e.g., Hai Luo AI, Yao Wen) and companion apps (e.g., Xing Ye, Mao Pao Ya).

Moon of the Dark Side’s companion app Xing Ye, despite high monthly active users, is viewed by MiniMax’s founder as a content‑community platform rather than a direct ChatGPT competitor, aiming to evolve into a Reddit‑ or Bilibili‑style ecosystem.

Zero‑One, led by Kai‑Fu Lee, rejects B‑side markets, focusing on C‑side products and overseas expansion with the productivity tool PopAi, while Zhipu and Baichuan concentrate on B‑side solutions, leveraging paid enterprise demand and sector‑specific applications such as healthcare.

Marketing spend has intensified: AI firms flood platforms like Bilibili with costly user‑acquisition campaigns (approximately ¥30 per new user for Kimi), and even place offline ads in subways. Although these campaigns boost traffic dramatically, they often do not translate into immediate revenue, leaving firms reliant on continuous financing.

Financing remains robust; all six companies raised over a hundred million yuan this year, with Baichuan completing a ¥5 billion A‑round and Zero‑One securing a multi‑dollar series round. Globally, AI startups attracted $18.3 billion in Q2 2024, accounting for 28 % of total VC funding, with notable contributions from Elon Musk’s xAI.

Personnel turbulence is evident: key executives from Zero‑One have departed for ByteDance and Beike, yet hiring for technical roles continues across the sector, indicating ongoing expansion plans despite internal churn.

Long‑term prospects hinge on achieving AGI and potential IPOs. OpenAI is rumored to pursue a public listing, while Chinese founders like Kai‑Fu Lee have expressed similar ambitions, acknowledging that a successful exit may require years of sustained investment.