On April 13 local time, a newly released annual Artificial Intelligence Index report from Stanford University revealed that China has completely closed the gap with the United States in top-tier AI model performance, marking a formal shift in the competition between the two nations from "the U.S. leading" to a new phase of "neck-and-neck racing."
The report, titled Artificial Intelligence Index 2026, released by Stanford University, is one of the most authoritative annual "health checkups" in the AI field. Its core conclusion is that the performance gap between China and the U.S. in top-tier AI models has essentially disappeared, signaling a global AI competition that has officially entered a new stage of "Sino-U.S. co-leading."
Data shows that as of March 2026, top American models led Chinese counterparts by only 2.7 percentage points. In 2023, this gap had been over 300 points.
Alternating Leadership: Since early 2025, both sides have repeatedly alternated at the top of rankings, with DeepSeek-R1 even achieving its first-ever tie with U.S. models.
First-Tier Contenders: On the human-judged leaderboard, scores of Chinese and American models have converged closely, with Anthropic, xAI, Google, OpenAI, Alibaba, and DeepSeek all ranking among the top tier.
This report serves as a sobering "intelligence briefing": while the U.S. still leads in foundational innovation and capital, China holds clear advantages in engineering implementation and scale. This implies that future competition between China and the U.S. in AI will fully shift from "can it be built?" to "can it be used well, affordably, and securely?"
Original source: toutiao.com/article/1862445221048320/
Disclaimer: The views expressed in this article are solely those of the author.