Chinese AI Models Generate 10% of OpenAI-Anthropic Revenue

Chinese artificial intelligence companies are seeing explosive user adoption, but their revenue remains a fraction of their American counterparts, creating a disconnect between market valuations and financial performance.

All Chinese AI models combined generate approximately 10% of the annual recurring revenue reported by OpenAI and Anthropic alone, according to estimates published by U.S.-based research firm Rhodium Group. The analysis, first reported by CNBC, examined industry metrics that project annual performance by multiplying recent monthly figures by 12.

Revenue breakdown reveals wide gap

Among Chinese AI developers, DeepSeek reported the lowest annual recurring revenue at $500 million, followed by MiniMax at $800 million and Moonshot at $1 billion. Z.ai disclosed $1.8 billion in its latest investor update. When including ByteDance's $4 billion and Alibaba's $2.4 billion, the combined Chinese total still falls dramatically short of OpenAI's $40 billion alone—before accounting for Anthropic's $65 billion.

The revenue shortfall becomes more pronounced when examining valuation multiples. Rhodium analysts calculated that Moonshot trades at roughly 50 times revenue, while DeepSeek commands an extraordinary 163 times multiple. These ratios significantly exceed OpenAI's 34x and Anthropic's 21x valuations.

Why it matters

The revenue-valuation mismatch threatens the sustainability of Chinese frontier AI labs at a critical growth phase. Unlike their U.S. competitors, Chinese developers face a narrower path to profitability while competing on price—leading Chinese models cost far less per task than OpenAI or Anthropic offerings, according to AI-comparison firm Artificial Analysis. This pricing pressure, combined with limited access to international capital markets and the open-source nature of many Chinese models, creates structural challenges for monetization even as user numbers climb.

Funding and market dynamics

Logan Wright, partner at Rhodium Group and report co-author, noted that the financing gap will make sustainable scaling difficult for Chinese AI labs. "They will be heavily dependent upon a favorable climate in the equity market—historically that's not an easy bet in China," Wright told CNBC.

Government support has concentrated on hardware infrastructure, with Rhodium estimating that state-affiliated sources provided over 60% of equity investment in Chinese AI chips and servers. However, direct funding for frontier labs appears less certain.

Several Chinese AI companies are pursuing public listings to address capital needs. Moonshot has reportedly filed confidentially for a Hong Kong IPO, while DeepSeek is also preparing for a potential listing. Anthropic is expected to list in the U.S. next month, with OpenAI targeting next year.

Z.ai raised its year-end annual recurring revenue forecast to $3 billion from $2.4 billion, though this remains well below U.S. competitors. The company's Hong Kong-traded shares have retreated to spring levels after briefly tripling over the summer.

Rhodium researchers noted that Chinese labs are exploring ways to capture more revenue from third parties that offer access to their models, though the open-source architecture of many Chinese systems allows independent deployment by anyone with adequate hardware.

The analysis was published by CNBC, with reporting by Evelyn Cheng.