Hugging Face’s latest observation report says the global open-source AI ecosystem is expanding quickly, with Chinese models emerging as a major force. According to the report, in most months since 2026, the largest open-source models released by Chinese AI companies have exceeded their U.S. counterparts by a wide margin in parameter count. The monthly upper bound for Chinese models ranged from 754 billion to 2.78 trillion parameters, while U.S. models stayed below 130 billion in most comparable months.
The report also says China’s edge is not limited to model size. Among Chinese models with more than 20 billion parameters, over 80% use relatively permissive licenses that allow secondary development and commercial deployment. Qwen has become a key foundation model within the Hugging Face open-source ecosystem, with more than 150,000 derivative models built on it, surpassing rivals such as Google and Meta. The report adds that, this year, the two U.S. companies releasing the most new open-source models are no longer Google and Meta, but AMD and Nvidia. Nvidia’s Nemotron series is being used to showcase GPU hardware coordination, while AMD has focused more on model conversion and performance tuning.
Hugging Face said in its latest observation report that the global open-source AI ecosystem is expanding rapidly, with Chinese models taking on a bigger role.
The report says that in most months since 2026, the largest open-source models released by Chinese AI companies have been significantly larger than those from U.S. peers in parameter count. The monthly upper bound for Chinese models ranged from 754 billion to 2.78 trillion parameters, while U.S. models were below 130 billion in most of those same months.
Hugging Face also said China’s strength is showing up in ecosystem development, not just model scale. Data from the platform shows that among Chinese models with more than 20 billion parameters, over 80% use relatively permissive licenses, allowing companies and developers to build on them and deploy them commercially.
Qwen has become an important foundation model in the Hugging Face open-source ecosystem, according to the report. Data shows that the number of derivative models built on Qwen has exceeded 150,000, well ahead of competitors including Google and Meta.
The report also notes that the two U.S. companies that have released the most new open-source models this year are no longer Google and Meta, but AMD and Nvidia. Nvidia launched its Nemotron series to show GPU hardware coordination through open models, while AMD has been more involved in model conversion and performance tuning.
The item was cited by Odaily, referencing China Media Group’s CCTV Finance.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
