A recent report by Rhodium Group highlights that all Chinese AI models combined generate only about 10% of the revenue reported by leading U.S. firms OpenAI and Anthropic. For instance, while OpenAI generates approximately $40 billion and Anthropic around $65 billion, major Chinese players like Z.ai, ByteDance, and Alibaba collectively report revenues far below these figures.
Z.ai's annual recurring revenue (ARR) is projected to reach $3 billion by year-end, but even this is dwarfed by the valuations of Chinese startups, with Moonshot and DeepSeek showing estimated revenue multiples of 50x and 163x, respectively. This contrasts sharply with OpenAI's 34x and Anthropic's 21x.
The report suggests that the high valuations of these Chinese firms may not be sustainable given their low revenue generation. Furthermore, the financing landscape poses challenges for these companies, as they rely heavily on favorable equity market conditions and government funding, which may not be guaranteed.
The volatility in the stock performance of Chinese AI companies, such as Z.ai and Minimax, further underscores the uncertainty in this sector. As the market evolves, the ability of Chinese AI labs to scale sustainably will be critical, especially as they explore new revenue models and face competition from more established U.S. firms