Chinese tech firm Meituan launched a new artificial intelligence model on Tuesday that it said was the first of its size to be trained using domestically developed computer chips.
China is locked in an intense race with the United States for AI supremacy, as Washington restricts exports of the most advanced semiconductor chips made by industry leader Nvidia.
In response to US restrictions imposed on national security grounds, Beijing has accelerated efforts to develop advanced domestic chips and reduce its reliance on American hardware.
Chinese technology giant Meituan on Tuesday unveiled LongCat-2.0, a new large language model (LLM) that it said performs on par with Google’s Gemini 3.1 Pro, released in February 2026.
LLMs underpin chatbots and many other AI applications. Generally, the greater the number of internal variables, or parameters, a model has, the more capable it is of handling complex tasks.
Trained on domestically produced chips
Meituan said LongCat-2.0 is the industry’s first AI model with more than one trillion parameters to be fully trained and operated on a cluster of 50,000 domestically produced chips.
The announcement marks a milestone for China’s rapidly developing AI sector, as training advanced models on vast datasets requires significant computing power.
