Alibaba's open-source Qwen models have surpassed the 3 billion download mark.
Chinese tech giant Alibaba has reached an impressive milestone: the cumulative number of downloads of their open-source AI models from the Qwen family over the past six months has exceeded 3 billion. This landmark event underscores a rapid shift in the global balance of power in the artificial intelligence market.
To date, the corporation has provided public access to more than 460 neural networks, on the basis of which third-party developers have created about 300,000 derivative models. Such an ecosystem demonstrates not just popularity, but the formation of a sustainable infrastructure around Qwen technologies.
Ecosystem Analytics: Hugging Face Data
My analysis of data from the Hugging Face platform, published in the semi-annual report, reveals an even more detailed picture. Over seven months of the current year, Qwen models were downloaded 2.05 billion times, and the number of derivative repositories reached 151,448. For comparison, the similar figure for Google is only 82,506. It is important to note that Hugging Face statistics do not include API requests and private deployments, making the real numbers even more impressive.
The local deployment segment is particularly telling. GGUF builds of Qwen are downloaded on average 39.6 million times per month, which is almost twice the figures for Gemma (20.8 million) and five times that of Llama (7.5 million). This indicates that Qwen has become the de facto standard for developers who need flexibility and control over models.
Success Factors and the Balance of Power
Qwen's success is explained by three key factors: regular updates, coverage of all scales—from sub-billion versions to the flagship Qwen3.8-Max with 2.4 trillion parameters—and the Apache 2.0 license, which removes all barriers to commercial use. The breadth of the lineup proved decisive: models with fewer than 1 billion parameters account for 83% of all downloads in the platform's history, while giants over 100 billion account for only 1%.
The report also records a tectonic shift: almost every month this year, the largest open-source model from China surpassed all American releases in size, reaching a ceiling of 754 billion to 2.78 trillion parameters. Meanwhile, in the United States, 41% of large models are released under proprietary licenses, which hinders their distribution. Notably, the leaders in the number of new open-source models in the US were not AI labs, but chip manufacturers—AMD and Nvidia.
My comment: These data confirm that Alibaba's openness strategy is not philanthropy, but a precise market calculation. By controlling the de facto standard for developers, the Chinese corporation is laying the foundation for dominance in the next generation of AI applications, where the key role will be played not by model size, but by its integration into real-world workflows.