Leading Chinese AI models like Kimi appear to be primarily trained on the outputs of US models (a process called distillation) rather than being built from scratch. This suggests China's progress is constrained by its ability to scrape and fine-tune American APIs, indicating the U.S. still holds a significant architectural and innovation advantage in foundational AI.
Related Insights
China is gaining an efficiency edge in AI by using "distillation"—training smaller, cheaper models from larger ones. This "train the trainer" approach is much faster and challenges the capital-intensive US strategy, highlighting how inefficient and "bloated" current Western foundational models are.
Challenging the narrative of pure technological competition, Jensen Huang points out that American AI labs and startups significantly benefited from Chinese open-source contributions like the DeepSeek model. This highlights the global, interconnected nature of AI research, where progress in one nation directly aids others.
The rise of Chinese AI models like DeepSeek and Kimmy in 2025 was driven by the startup and developer communities, not large enterprises. This bottom-up adoption pattern is reshaping the open-source landscape, creating a new competitive dynamic where nimble startups are leveraging these models long before they are vetted by corporate buyers.
Despite strong benchmark scores, top Chinese AI models (from ZAI, Kimi, DeepSeek) are "nowhere close" to US models like Claude or Gemini on complex, real-world vision tasks, such as accurately reading a messy scanned document. This suggests benchmarks don't capture a significant real-world performance gap.
Chinese AI models like Kimi achieve dramatic cost reductions through specific architectural choices, not just scale. Using a "mixture of experts" design, they only utilize a fraction of their total parameters for any given task, making them far more efficient to run than the "dense" models common in the West.
Despite leading in frontier models and hardware, the US is falling behind in the crucial open-source AI space. Practitioners like Sourcegraph's CTO find that Chinese open-weight models are superior for building AI agents, creating a growing dependency for application builders.
America's competitive AI advantage over China is not uniform. While the lead in AI models is narrow (approx. 6 months), it widens significantly at lower levels of the tech stack—to about two years for chips and as much as five years for the critical semiconductor manufacturing equipment.
According to DeepMind CEO Demis Hassabis, while Chinese AI models are rapidly closing the capability gap with US counterparts, they have yet to demonstrate the ability to create truly novel breakthroughs, like a new transformer architecture. Their strength lies in catching up to the frontier, not pushing beyond it.
While the U.S. leads in closed, proprietary AI models like OpenAI's, Chinese companies now dominate the leaderboards for open-source models. Because they are cheaper and easier to deploy, these Chinese models are seeing rapid global uptake, challenging the U.S.'s perceived lead in AI through wider diffusion and application.
To escape platform risk and high API costs, startups are building their own AI models. The strategy involves taking powerful, state-subsidized open-source models from China and fine-tuning them for specific use cases, creating a competitive alternative to relying on APIs from OpenAI or Anthropic.
