The US technology sector has continued its decline this week. While news of a potential OpenAI IPO delay served as the immediate trigger, analysts point to Zhipu's low-cost AI model as perhaps the more profound threat—impacting not just enterprise clients of OpenAI and Anthropic but also shaking optimistic market expectations for AI infrastructure spending.
According to Jefferies strategist Christopher Wood, citing industry sources, the GLM5.2 model launched by Zhipu "performs almost as well as Anthropic's, but at a quarter of the cost per token." He characterized the past week as "another DeepSeek moment." The analysis suggests that low-cost AI models will influence enterprise customer choices and undermine the investment thesis for AI infrastructure.
During Friday's US trading session, Micron Technology shares fell over 7%, while Advanced Micro Devices Inc (AMD) and Intel Corporation (INTC) both dropped more than 4%. Oracle Corporation (ORCL) hit a new low after declining 19% over the past five trading days.
Zhipu's Low-Cost Model Disrupts Frontier AI Pricing
Zhipu's GLM5.2 model is seen as a core variable in the tech sector's recent pressure. A Jefferies analyst report noted the model not only closely matches the performance of Anthropic's flagship product but also offers privacy protection levels comparable to leading models.
A Morgan Stanley trader stated that Zhipu's new model demonstrates "extremely strong programming capabilities" and represents direct competition in the enterprise market. The trader noted that the trend of enterprises and hyperscale cloud providers shifting to lower-cost models resembles "a recalibration of willingness-to-pay for AI, rather than a decline in AI demand."
Deutsche Bank analyst Jim Reid highlighted similar pressures in a client report, noting that DeepSeek's V4-Pro model performs "similarly to Anthropic's flagship Claude Fable 5 on about 90% of daily tasks, but at about 1.5% of the cost."
Enterprise AI Spending Landscape Rapidly Reshaping
AI pricing pressures have triggered a chain reaction on the enterprise side. In recent months, several companies have begun scaling back AI expenditures or exceeding original budgets, with token pricing for high-end AI models becoming a significant pain point.
Jefferies analysts further noted that if low-cost models also offer reliable privacy protection, enterprises have an incentive to move AI workloads from cloud providers back to their own servers. This would fundamentally alter the investment logic for AI infrastructure.
"Demand structure is clearly tilting towards low-cost models," wrote a Morgan Stanley trader. This trend poses valuation pressure for stocks across the NVIDIA Corporation (NVDA) supply chain, cloud computing platforms, and data center construction.
OpenAI and Anthropic IPO Prospects Dim
Rumors of an OpenAI IPO delay, combined with intensifying competition, are making the market more cautious about the valuation expectations for these leading AI companies.
Analysts worry that if enterprise customers continue migrating to lower-priced models, the revenue growth potential of OpenAI and Anthropic ahead of their IPOs could be eroded. A potential price war between them could further compress their respective valuation ranges before going public.
The continued emergence of cheaper, performance-competitive open-source AI models is seen as a more profound threat to these companies' proposed IPOs and the broader tech sector. The market's previous assumption of an "astronomical growth trajectory" for AI infrastructure spending is now facing reassessment pressure.
Comments