US Lead in the AI Race With China Is Rapidly Narrowing

A Bloomberg graphics feature, produced jointly with Vals.ai, tested seven frontier Chinese and US AI models on a real-world e-commerce task. According to the excerpt, most of the models scored 100% functional accuracy. The models had very different price tags. The piece is titled "US Lead in the AI Race With China Is Rapidly Narrowing." The feature is a visual presentation from Bloomberg's graphics section, and the excerpt does not name the specific models involved.

The test's defining result is that practical performance converged while pricing diverged. Across the seven frontier models, full functional accuracy was the norm, meaning the systems achieved the same successful outcome on the assigned task. The term "functional accuracy" is used in the excerpt to describe the models' performance on this task. The principal measurable distinction between the Chinese and US entries was cost. The article does not specify which models were cheaper, nor does it list exact prices in the excerpt; it states only that price tags were very different. The excerpt offers no further benchmark scores or error rates.

The article suggests the US lead is narrowing because Chinese models now achieve comparable functional results, with pricing acting as a differentiating factor. The findings come from a joint Bloomberg and Vals.ai evaluation, and the article frames the e-commerce task as a real-world test rather than a synthetic benchmark. No further details about the methodology are included in the excerpt.