Chinese corporation Alibaba has announced an impressive milestone: the total number of downloads of its open AI models from the Qwen family has exceeded 3 billion in just the last six months. This is not just a number, but a marker of a fundamental shift in the industry, where open weights are becoming the main weapon in the competitive struggle.
To date, Alibaba has provided the community with more than 460 neural networks, on the basis of which developers around the world have created over 300,000 derivative models. The scale is impressive, but even more interesting is the dynamics recorded by independent platforms.
Hugging Face Analytics: Dominance Statistics
The independent platform Hugging Face, in its latest report on the state of open models, confirms this trend. According to the platform's calculations, in just the first seven months of this year, Qwen models were downloaded 2.05 billion times, and the number of forks (derivative repositories) reached 151,448. For comparison, the closest competitor Google has only 82,506.
The gap with competitors is colossal. Over the same period, Google model downloads are estimated at 418 million, and Meta at 227 million. It is important to understand that Hugging Face statistics only account for activity within its own ecosystem, ignoring API requests, private corporate deployments, and other distribution channels. Thus, the real scale of Qwen's penetration could be significantly higher.
The report's authors rightly caution against a straightforward interpretation of these figures as market share. However, they cannot be ignored. The number of Qwen derivative repositories grows daily by 180–210, indicating the deepest integration of the model into the workflows of developers worldwide. Also telling is the fact that out of 28,531 GGUF conversions (a format for local execution), only 54 were created by Alibaba itself — the rest is the work of the community, which actively adapts the model to its tasks.
Success Factors: Strategy, Reach, and License
Why exactly has Qwen managed to outpace giants like Google and Meta? I highlight three key factors. First, the regularity of updates and the breadth of the model lineup — from compact versions with sub-billion parameter counts to the flagship Qwen3.8-Max with 2.4 trillion parameters. Second, the flexible Apache 2.0 license, which removes all restrictions on commercial use and modification. Third, and perhaps the decisive factor, is the focus on small models. According to Hugging Face data, versions 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%.
This explains why laboratories focused exclusively on large LLMs are falling behind. For example, Moonshot AI, which releases almost no models smaller than 70 billion, has gathered only 37 million downloads in a year — roughly 55 times fewer than Qwen. In the local deployment segment (GGUF), Qwen also leads with 39.6 million monthly downloads versus 20.8 million for Gemma and 7.5 million for Llama.
Geopolitics of Open Source
The report's data also reveals a shift in the balance of power. The largest open model from China this year almost every month surpassed all American releases in size, reaching a ceiling of 2.78 trillion parameters. Moreover, in China, 59% of models with more than 20 billion parameters are released under Apache 2.0, and none have restrictions on commercial use. In the US, the picture is different: 41% of such models are under proprietary licenses, which hinders their distribution.
Notably, in the US, the leaders in the number of new open models were not AI laboratories but chip manufacturers — AMD and Nvidia, each releasing more than 200 repositories. This suggests that infrastructure players are betting on the open ecosystem, while model developers try to maintain control.
My analysis: Qwen's success is not just a victory for one company, but a demonstration that an open model with the right scaling and licensing strategy can capture the real developer market faster than closed or semi-open counterparts. While American giants are torn between commercial gain and openness, China is methodically occupying the niche of the de facto standard for engineers worldwide.