Key facts
- Chinese AI models pose a key risk to Anthropic ahead of its IPO, according to Hedgeye Risk Management.
- Anthropic's third-quarter revenue rose nearly 50% to $16.8 billion, marking a slowdown in growth.
- Chinese open-weight models now serve about half of all AI tokens globally.
- Chinese AI models are cheaper than US counterparts while trailing slightly in capabilities.
- Anthropic's IPO valuation target has reportedly decreased to around $1.5 trillion.
- Anthropic claims significant progress in curbing illicit use of its model outputs by Chinese firms.
The growing popularity of Chinese artificial intelligence models presents a significant risk for Anthropic as it prepares for its initial public offering, according to Hedgeye Risk Management. Felix Wang, a managing director at Hedgeye, noted that clients are increasingly inquiring about Chinese competition.
Anthropic reported a nearly 50% increase in third-quarter revenue to $16.8 billion, a slowdown compared to earlier in the year. This deceleration coincides with a rise in the global popularity of Chinese open-weight AI models, which now account for about half of all AI tokens used worldwide. Companies like Z.ai and DeepSeek are experiencing rapid revenue growth and increasing prices for their AI models, which are noted for being cost-effective alternatives to US offerings like Anthropic's Claude.
While Chinese models lag US counterparts in capabilities by only a few months, they are significantly cheaper, creating pricing pressure. This competitive landscape, coupled with a cooling of enterprise AI spending and Anthropic's efforts to curb the illicit use of its model outputs by Chinese firms, has contributed to the revenue slowdown. Anthropic's IPO is reportedly expected next month, with investor discussions centering on a valuation around $1.5 trillion, a reduction from its previous target.