FrontierThe story, in brief
China’s Meituan open-sources massive LongCat-2.0 AI model, saying it was trained on domestic chips
1.6 trillion parameters. Trained entirely on Chinese chips. Meituan just showed the world domestic AI infrastructure actually works.
The KeyNews take
Why it matters
China's supply-chain independence in AI is no longer theoretical—a major tech company just proved large-scale model training is viable on domestic silicon, challenging the Western GPU monopoly and reshaping geopolitical AI competition.
The key facts
5 to knowLongCat-2.0 model: 1.6 trillion parameters
Trained on domestic Chinese chips and compute clusters
Open-source release signals confidence in domestic hardware stack
Meituan (food delivery/super-app giant) expanding into frontier AI
Implications for US semiconductor export restrictions and Chinese AI independence
Go to the source
SiliconAnglesiliconangle.com
Publisher excerpt: Beijing, China-based Meituan Inc. today debuted its next-gen LongCat-2.0 open-source large language model, stating that the company trained the 1.6-trillion-parameter model on domestic Chinese chips and compute clusters. The larger takeaway for this colossal model isn’t just the open-source…