Leading Chinese artificial intelligence firms are intensifying the pressure on Silicon Valley. Moonshot AI and Alibaba have unveiled models they say can compete with the top offerings from OpenAI and Anthropic at a fraction of the cost. The rapid cadence of releases signals waning U.S. leadership in cutting-edge AI as technology plays a pivotal role in national security, economic power, and geopolitical clout.
The Beijing-based Moonshot AI on Friday unveiled the Kimi K3 model. In internal testing, the system reportedly outperforms nearly all American rivals, trailing only OpenAI’s GPT-5.6 Sol and Anthropic’s Claude Fable 5, though on some standardized benchmarks it even exceeds them.
Openness as a competitive strategy
This weekend, the Chinese technology giant Alibaba previewed the Qwen 3.8 model, which it calls one of the most powerful available today and second only to Claude Fable 5 — Anthropic’s flagship system. Both companies emphasize a key difference from leading U.S. laboratories: instead of keeping their most advanced models closed, they are opening them up for wide use.
Although some American companies, including Meta, also adopt a similar approach, the release of models that developers can freely download and modify is becoming a major competitive advantage for China’s AI industry. Moonshot AI calls Kimi K3 the largest open AI model in the world with 2.8 trillion parameters. Alibaba reports that Qwen 3.8 has 2.4 trillion parameters and is still under improvement. At the same time, the companies note that a larger number of parameters does not always translate to better model quality, and OpenAI and Anthropic do not disclose these indicators at all.
Technological race and export restrictions
It is difficult to assess the real capabilities of Chinese models until their full release and independent testing. Moonshot AI plans to publish the full set of trained model parameters (model weights) on July 27, and Alibaba says it will open the parameters of Qwen 3.8 soon. These releases have heightened the competition between the U.S. and China in what is increasingly being called the defining technological race of modern times.
The new models also call into question whether massive investments by American companies in chips and data centers will provide a long-term advantage if Chinese competitors can achieve similar results at lower costs. Such openness from China comes amid Washington’s steps to restrict global access to basic technologies, including export controls on advanced chips and pressure on Anthropic not to release its most powerful system due to fears it could help foreign competitors catch up with the United States faster.
Similar to how DeepSeek last year presented a model capable of competing with American systems at a fraction of their cost, the new developments are likely to impact economic policy and national security policy. They demonstrate that the U.S. technological lead may be significantly smaller than previously thought.
Source: The Verge
Meta: The rapid release of new systems indicates that U.S.
