Newsletter Subscribe
Enter your email address below and subscribe to our newsletter
[forminator_form id="25163"]

tech.yahoo+1cnbccnbc+1Werner Vogels, Amazon's Amazon.com, Inc. chief technology officer, said companies concerned about escalating artificial intelligence expenses are increasingly turning to open-source models as a cost-effective alternative to proprietary systems, a shift being accelerated by U.S. government restrictions on advanced AI releases.tech.yahoo
Vogels made the remarks at the UN's AI for Good Summit in Geneva, which ran from July 7 to 10, where he appeared alongside other tech executives to discuss the future of AI development and governance.fortune+1
The trend Vogels identified aligns with broader industry data showing that open-source and Chinese-built AI models can be "60% to 90% cheaper" than leading proprietary systems from Anthropic and OpenAI, according to Justin Summerville at OpenRouter, who spoke with CNBC. A Citi research note found that Chinese models charge as little as 18 cents per million tokens, compared with an average of $4 for top proprietary models.finance-commerce+1
Chinese AI models are now estimated to be only six to nine months behind leading U.S. frontier systems in capability, according to Kyle Chan, a fellow at the Brookings Institution's John L. Thornton China Center. "The new open source models are performing well and prove capable for all but the most complex LLM tasks," Summerville told CNBC.cnbc
The open-source momentum is perhaps best illustrated by Alibaba's Qwen model family, which surpassed one billion cumulative downloads on Hugging Face in early 2026, overtaking Meta's Llama to become the world's most downloaded open-source AI framework. Qwen now averages approximately 1.1 million downloads per day and has spawned more than 200,000 derivative models. According to the New York Times, Qwen became the world's most downloaded open-source AI system by January.nytimes+2
Researchers at Interconnects AI found that Qwen generated 153.6 million downloads in February alone — more than double the combined total from the next eight major players, including Meta, DeepSeek, and OpenAI.scmp
The shift toward open-weight models has been reinforced by recent U.S. government actions restricting proprietary AI releases. Last month, the Department of Commerce forced Anthropic to pull its Fable 5 and Mythos 5 models offline over national security concerns, a move that The Hill reported has spurred demand for open-source alternatives. U.S. House committees launched a joint investigation in April into growing enterprise adoption of Chinese AI models, even as companies continue gravitating toward the cheaper options available outside the proprietary ecosystem.thehill+2