China’s open-source artificial intelligence (AI) models are expanding their influence across the world, creating a new competitive challenge for US technology giants.

A Bloomberg report revealed that Chinese models are now outperforming US counterparts on the popular developer platform Hugging Face. Data shows that Alibaba’s “Qwen” models have recorded around 385.3 million downloads, overtaking Meta’s “Llama” models, which stand at 346.2 million.

Currently, over 40 percent of new language models on Hugging Face come from Chinese developers. In contrast, Meta’s share has dropped to around 15 percent, indicating a rapid shift in developer preference toward Chinese tools.

Growing global interest in Chinese AI

Industry experts say that this momentum is already shaping the strategies of major global firms. Venture capitalist Chamath Palihapitiya said on his All-In podcast that his company has shifted major AI workloads to the “Kimi K2” model developed by Moonshot AI, citing lower costs and efficiency.

Similarly, Airbnb CEO Brian Chesky acknowledged that his company skipped integration with OpenAI’s ChatGPT because its connectivity tools “weren’t quite ready.” Instead, Airbnb is using Alibaba’s Qwen models, which Chesky described as “very good. It’s also fast and cheap.”

US retains hardware strength but faces rising competition

The United States still leads in AI hardware and chip technology, largely due to companies like Nvidia Corp. Nvidia CEO Jensen Huang has repeatedly stated that America must “race ahead and win developers worldwide,” even while admitting that China is “just nanoseconds behind.”

However, analysts say that for many startups and small companies, open-source Chinese models offer more advantages. They are easier to access, cheaper to run, and simpler to modify for local needs. These factors make them attractive alternatives to proprietary US systems that often come with restrictions and higher costs.

A recent study by a global open-source AI coalition found that developers are increasingly choosing Chinese models because of their flexibility, cost-effectiveness, and ease of customization.

Strategic concerns and shifting priorities

Experts warn that this growing dependence on foreign AI systems could raise data governance and security concerns. But for many companies focused on staying competitive, affordability and performance are taking priority over geopolitical risks.

Although it is still early to predict the final outcome of the global AI race, the pattern is clear. Chinese models are capturing developer attention faster than expected, and this trend could reshape the future of the AI industry.

As one expert told Bloomberg, the critical question now is not “who has the best hardware?” but “which ecosystem are developers building on?”