According to estimates from U.S.-based research firm Rhodium Group, OpenAI and Anthropic collectively generate approximately ten times the annual recurring revenue (ARR) of all major Chinese AI models combined. The report highlights a significant disparity in monetization, with OpenAI reporting $40 billion in ARR and Anthropic at $65 billion, compared to lower figures for Chinese competitors such as Z.ai ($1.8 billion), ByteDance ($4 billion), and Alibaba ($2.4 billion). DeepSeek recorded the lowest ARR among major Chinese firms at $500 million, followed by MiniMax at $800 million and Moonshot at $1 billion.
Despite rapid adoption rates, the low revenue generation has raised concerns regarding company valuations. Rhodium noted that valuation-to-revenue ratios for Moonshot and DeepSeek appear exorbitant at estimated multiples of 50x and 163x, respectively, far exceeding the 34x for OpenAI and 21x for Anthropic. While some Chinese firms like Z.ai project their ARR will reach $3 billion by year-end, the financing gap remains a critical challenge. Logan Wright, partner at Rhodium Group, stated that Chinese frontier labs face difficulties scaling sustainably due to heavy dependence on favorable equity market conditions and limited direct government funding for software development.
The stark revenue divergence underscores a fundamental structural difference between the U.S. and Chinese AI ecosystems. While Chinese models benefit from widespread adoption and open-source accessibility, they struggle to convert usage into substantial recurring revenue compared to their closed-source American counterparts. This dynamic suggests that the current valuation premiums placed on Chinese startups may be disconnected from their actual cash-generating capabilities, creating potential fragility in the sector's financial health. The reliance on high valuation multiples rather than robust earnings indicates that investor sentiment is driving prices more than operational fundamentals.
From an institutional perspective, the sustainability of Chinese frontier AI labs appears heavily contingent on external capital flows rather than organic growth. With state-affiliated sources providing over 60% of equity investment in hardware infrastructure but likely hesitating to fund software labs directly, these companies face a precarious financing environment. The upcoming IPO plans for firms like Moonshot and DeepSeek will serve as a critical test of whether public markets are willing to absorb these high valuations despite the evident revenue gaps. Investors should watch for signs of correction in secondary market trading, as seen in recent volatility for listed entities like Z.ai and MiniMax, which may signal broader reassessment of the sector's risk profile.
