Chinese technology companies are rapidly advancing in artificial intelligence (AI), closing the gap with their U.S. counterparts despite facing restrictions on advanced chip access. Chinese firms such as Huawei, ByteDance, Alibaba, Baidu, and Tencent are developing competitive machine learning models, with some reports indicating that Tencent’s MoE model matches Meta’s Llama 3.1 while utilizing fewer resources. The Chinese model DeepSeek v3 has demonstrated impressive capabilities, requiring only 2.788 million H800 GPU hours for its full training. Analysts suggest that the U.S. export controls on chips may not be achieving their intended effects, as Chinese companies innovate under resource constraints, leading to breakthroughs in AI development. This shift indicates a growing competitiveness in the global AI landscape, with Chinese firms leveraging open-source models and creative training methods to enhance their technological prowess.