@aaditsh: China gave its models away when it was behind. Qwen, DeepSeek, GLM. All were open. The strategy was to commoditize the …
Summary
Analysis of China's open-source AI strategy, which initially gave away models to commoditize the layer, leading to significant token routing through Chinese models. Now, as Chinese models approach frontier capabilities, Beijing is considering restricting access.
View Cached Full Text
Cached at: 07/07/26, 09:38 PM
China gave its models away when it was behind.
Qwen, DeepSeek, GLM. All were open. The strategy was to commoditize the model layer so US labs can’t charge premium prices.
It worked. US companies now route over 30% of their AI tokens through Chinese models every week (a year ago that was 11%).
Now Chinese models are roughly 6 to 9 months behind the US frontier instead of years. And this week it’s reported that Beijing is discussing restricting overseas access to its best models for the first time.
The logic is pretty simple. You give the model away when you’re second. You restrict it when you think you’re about to be first.
Ethan Mollick (@emollick): This is a key reason I don’t expect the flow of frontier open weights models to continue indefinitely, or even for very much longer.
Similar Articles
@francoisfleuret: Why are Chinese companies releasing very strong open-source models?
A discussion question about the trend of Chinese companies releasing strong open-source AI models.
@rohanpaul_ai: This is bad news. Now, frontier open source access also will probably no more be available for all. China is preparing …
China is preparing to limit foreign access to its strongest AI models, a move that could raise global AI costs and split the model market by nationality, with Beijing holding talks with Alibaba, ByteDance, and Z.ai about keeping advanced models domestic.
@rohanpaul_ai: A paper on open model adoption and finds Chinese models, led by Qwen, now dominate. China passed the U.S. in open model…
This paper analyzes open language model adoption, finding that Chinese models led by Qwen now dominate downloads, surpassing US models by March 2026. Qwen's lead comes from a diverse range of model sizes, while DeepSeek leads in very large models, and some US models still show strong momentum.
Will Chinese Open Source Models be the only option soon?
Discusses the potential for Chinese open source AI models to become the dominant option in the near future.
@svpino: Cina's AI ecosystem is different from everyone else: Every company wants to release its own SOTA model, but they are al…
The tweet discusses China's AI ecosystem where companies compete but release open-source models. It notes a mystery model 'Owl Alpha' on OpenRouter revealed to be Meituan's LongCat-2.0, a 1.6T parameter model with ~48B activated parameters optimized for reasoning and tool calling.