China’s AI Models Generate Just 10% of OpenAI and Anthropic’s Revenue, Report Finds

TLDR

• China’s top AI models combined earn roughly 10% of what OpenAI and Anthropic make in annual recurring revenue
• OpenAI’s ARR stands at $40 billion; Anthropic’s is $65 billion
• ByteDance leads Chinese AI firms with $4 billion ARR; DeepSeek sits at just $500 million
• Chinese AI startups carry high valuation-to-revenue multiples, with DeepSeek at 163x versus OpenAI’s 34x
• Financing constraints and equity market uncertainty could limit Chinese AI labs’ ability to scale

---

China’s AI companies are growing fast but still earn a small fraction of what their US rivals bring in, according to new estimates from research firm Rhodium Group.

JUST IN: OpenAI & Anthropic are generating roughly 10x more revenue than all Chinese AI models combined. — CNBC

— Polymarket (@Polymarket) September 17, 2026

The report found that all of China’s leading AI models combined generate about 10% of the annual recurring revenue reported by OpenAI and Anthropic. That gap is large and shows no signs of closing quickly.

Rhodium put OpenAI’s ARR at $40 billion and Anthropic’s at $65 billion. Those numbers dwarf what Chinese firms are pulling in right now.

ByteDance leads the Chinese pack with $4 billion in ARR, followed by Alibaba at $2.4 billion. Z.ai, formerly known as Zhipu AI, told investors this week its ARR had reached $1.8 billion.

Moonshot came in at $1 billion, MiniMax at $800 million, and DeepSeek at $500 million. Even added together, these figures fall well short of OpenAI alone.

Valuations Outpace Revenue

Despite lower revenues, Chinese AI startups are being valued at high multiples. Rhodium estimated DeepSeek’s valuation-to-revenue ratio at 163x and Moonshot’s at 50x.

By comparison, OpenAI sits at 34x and Anthropic at 21x. The Rhodium report called the valuations for Moonshot and DeepSeek “exorbitant” relative to their current revenue.

This means investors are betting heavily on future growth that has not yet shown up in the numbers.

Funding and Scaling Challenges

Logan Wright, a partner at Rhodium Group who co-authored the report, said Chinese AI labs face real funding challenges going forward.

“The financing gap means it will be far more difficult for Chinese frontier AI labs to scale sustainably,” Wright said. He added that state funding in China has mostly gone toward chips and servers, not model developers.

Rhodium estimated that state-affiliated sources made up more than 60% of equity investment in Chinese AI chips and hardware infrastructure.

Wright also noted the companies will rely heavily on equity markets, which have historically been unpredictable in China.

Data from Artificial Analysis shows that top US models from OpenAI and Anthropic remain mostly closed-source and cost more per task than Chinese alternatives. Chinese models have attracted users partly through lower prices.

Several Chinese AI companies are now pursuing public listings. Moonshot has reportedly filed confidentially for a Hong Kong IPO, and DeepSeek is also said to be preparing to list.

On the US side, Anthropic is expected to go public in October. OpenAI has reportedly pushed its own listing plans to next year.

The Rhodium report paints a clear picture: Chinese AI firms are growing, but the revenue gap with US leaders remains wide, and the path to closing it is uncertain.