Chinese AI Models Hit 46% Usage Peak Among U.S. Enterprises Since February

According to OpenRouter data, Chinese AI models' token usage share among U.S. enterprises reached a peak of 46% since early February, compared to an average of 11% over the prior 12 months and just 4.5% in the first half of 2025. The surge reflects enterprise cost sensitivity as OpenAI and Anthropic raise flagship model prices. OpenRouter analysis shows Chinese open-source models cost 60% to 90% less than leading Western alternatives. AI startup Lindy switched its full traffic from Anthropic's Claude to DeepSeek in June, with CEO Flo Crivello noting the cost curve "dropped straight to the floor," projecting savings of millions of dollars within months.

Zhipu's GLM 5.2, released in June, scores within 1 percentage point of Anthropic's Opus 4.8 on intelligent agent benchmarks while costing roughly one-fifth as much. Vercel's agent infrastructure lead Harpreet Arora reported GLM 5.2 achieved the fastest model adoption rate Vercel tracked this year, with daily token volumes growing approximately 27-fold in its first full week.