China’s AI startups can match the U.S.’s models. They can’t yet match the U.S.’s money

China’s artificial intelligence sector is moving closer to the United States in model development, according to estimates cited in the source material. The progress places leading Chinese systems within months of the latest releases from major U.S. AI companies.

The gap has tightened

Analysts estimate that China’s strongest AI models are now about four months behind the most sophisticated systems from OpenAI and Anthropic. That represents a notable reduction from the beginning of the year, when the estimated gap stood at seven months.

The figures suggest that the pace of Chinese AI development is accelerating relative to the frontier established by U.S. companies. While the estimate does not indicate that the two sides have reached parity, it points to a considerably narrower distance in model capabilities.

Moonshot’s Kimi K3 highlights China’s progress

Moonshot’s Kimi K3 is identified as the world’s largest open-weight model and has approached the performance of leading American frontier systems. Its progress is part of a broader shift in which Chinese AI developers are increasingly challenging the technical lead held by U.S. companies.

For the AI industry, the shrinking development gap increases the importance of comparing models by performance and release timing rather than by national origin alone. It also shows how quickly the competitive landscape is changing, even though the most advanced U.S. systems remain ahead by the cited estimate.

What comes next

The current trajectory makes the distance between Chinese and U.S. AI development a closely watched measure. For now, the available estimates show rapid progress in China and continued leadership by the latest U.S. frontier models.