The performance gap between China’s top AI models and their U.S. counterparts has narrowed to a record-low 3%. [3] At the start of the year, the gap was 15%. [15] New models from startups such as DeepSeek are driving the shift, with the company’s latest version now ranking among the world’s top models. [3, 12]
More than 1,100 large language models are competing in China’s crowded domestic market. [3] Tech giants Alibaba, Baidu, and Tencent are increasingly targeting enterprise AI applications. Well-funded startups including Moonshot AI and Z.ai are also releasing competitive models. [6, 15]
The rivalry has fueled intense price wars, creating profitability challenges across the sector. [12]