U.S. companies are increasingly adopting Chinese-built AI models from DeepSeek and Z.ai as cost pressures mount and performance gaps narrow with American rivals. Since Feb. 8, Chinese models have accounted for over 30% of weekly token usage by U.S. firms on OpenRouter, a developer platform, peaking at 46%—up from an 11% average over the previous 12 months. The shift comes as token prices for advanced models from OpenAI and Anthropic rise, with Chinese alternatives offering 60% to 90% cost savings while operating six to nine months behind top U.S. systems in capability, according to Brookings fellow Kyle Chan. At the end of June, the U.S. government requested OpenAI limit rollout of new models and lifted export controls on Anthropic's Mythos and Fable systems.
U.S. Companies Increase Chinese AI Model Usage on Cost Pressures
Companies deploying AI models for product development and internal efficiencies are experimenting with cheaper open-source and open-weight models, with the most capable systems produced by Chinese firms. Open-source and open-weight models allow developers to inspect, use, and sometimes modify AI systems, differing from closed proprietary models by OpenAI, Anthropic, and Google.
In June, AI startup Lindy moved 100% of its traffic from Anthropic's Claude models to DeepSeek. CEO Flo Crivello told CNBC the decision will save Lindy millions of dollars within months, stating the cost curve "crash to the ground" after the switch.
DeepSeek's share of gateway tokens climbed between May and June on Vercel, a platform for deploying apps and websites. Z.ai's GLM 5.2, released in June, saw the fastest adoption of any model tracked by Vercel in 2026. Harpreet Arora, head of agentic infrastructure at Vercel, told CNBC that in GLM 5.2's first full week after launch, daily token volume grew approximately 27 times and customer usage grew approximately 80 times.
"Price is doing the work here," Arora said. "When a task doesn't need the best model, teams are beginning to route it to the cheapest one that's good enough, and the recent wave of models coming out of China is winning that trade."
Open-source Chinese models can be "60% to 90% cheaper" than leading Anthropic and OpenAI models, according to Justin Summerville, who works on data and analytics at OpenRouter.
On LaunchLemonade, an AI agent platform for regulated industries, GLM 5.2 is now in the top five models by usage, with Claude and ChatGPT still dominating. CEO and founder Cien Solon told CNBC that Chinese models like Z.ai and Alibaba's Qwen are becoming options for companies as they offer an attractive combination of performance and cost for specific workloads.
Chinese AI Models Close Performance Gap with U.S. Rivals
Chinese AI models operate at a "fraction of the cost" of U.S. rivals while performing "close to the top American frontier models," Chan said, estimating they are currently "six to nine months" behind top U.S. systems.
"The new open source models are performing well and prove capable for all but the most complex LLM tasks," Summerville said.
GLM 5.2 landed within a percentage point of Anthropic's Opus 4.8 on one closely watched agentic benchmark at roughly a fifth of the cost. Some researchers have said GLM 5.2 can perform on par with top U.S. labs on some cyber benchmarks. Switching to DeepSeek V4 increased performance on many core use cases for Lindy, Crivello said in a post on X.
Yacine Jernite, head of machine learning at Hugging Face, told CNBC: "We're seeing companies increasingly motivated to turn to cheaper AI stacks they can control and adapt themselves, and given the state of open-source and open-weight models that often means leveraging Chinese options. There is a real risk that users get stuck having to choose between performant but expensive US proprietary models whose price and accessibility can quickly fluctuate, or using Chinese models as the only feasible alternative whenever they want to control costs or own their AI stack."
FAQ
What percentage of U.S. company token usage on OpenRouter comes from Chinese AI models?
Since Feb. 8, Chinese AI models have accounted for over 30% of weekly token usage by U.S. companies on OpenRouter, peaking at 46%. This compares to an 11% average over the previous 12 months and 4.5% in the first half of 2025.
How much cheaper are Chinese AI models compared to U.S. alternatives?
Open-source Chinese AI models are 60% to 90% cheaper than leading models from OpenAI and Anthropic, according to Justin Summerville of OpenRouter. For example, Z.ai's GLM 5.2 costs roughly one-fifth of Anthropic's Opus 4.8 while delivering comparable performance on certain benchmarks.
Which U.S. company switched entirely to Chinese AI models in June?
AI startup Lindy moved 100% of its traffic from Anthropic's Claude models to DeepSeek in June. CEO Flo Crivello stated the switch will save the company millions of dollars within months.
