Chinese-developed language models now account for roughly 61% of token consumption among the top 10 models on OpenRouter, the popular API routing platform. The majority of compute flowing through this key American AI infrastructure layer is being routed to models built in Beijing, Hangzhou, and Shenzhen.
In late 2024, Chinese open-weight models represented less than 1.2% of weekly token consumption. By April 2026, Chinese provider traffic had hit 51% of all tokens processed on the platform.
The leading names include Alibaba’s Qwen, DeepSeek, Moonshot AI’s Kimi, Zhipu AI’s GLM, and MiniMax. The share of programming and agentic workloads on OpenRouter climbed from about 11% to over 50% through 2025. American startups are building core products on these models, with code generation and autonomous agents increasingly handled by models developed outside the US.
The driving factors are straightforward: cost and capability. Chinese providers have priced APIs aggressively, often undercutting US counterparts. The open-weight nature of many models allows startups to inspect, fine-tune, and deploy with fewer restrictions.
For Chinese companies, particularly Alibaba, Qwen’s presence means integration into American software products. US export controls on advanced chips have spurred Chinese labs to focus on efficiency, producing competitive models at lower computational cost.