Alibaba's Qwen tops 3 billion downloads, overtaking Meta and Google

Alibaba's Qwen family has become the world's most-downloaded open-weight AI models, amassing more than 3 billion downloads in six months and overtaking Google's 418 million and Meta's 227 million, Hugging Face data show.

"Qwen has become part of the default workflow for developers deciding what models to fine-tune and deploy," the Hugging Face report said.

Alibaba has open-sourced more than 460 models through the Qwen family, spawning over 300,000 derivative models. Qwen derivative repositories grow by 180-210 daily, and of the 28,531 GGUF conversions on the platform, only 54 were created by Alibaba — the rest by the community. Models with fewer than 1 billion parameters account for 83% of all downloads in the platform's history, while neural networks above 100 billion account for just 1%.

The breadth of the model range proved decisive. Moonshot AI, which rarely releases models smaller than 70 billion parameters, gathered 37 million downloads in a year — about 55 times less than Qwen. Qwen's GGUF builds are downloaded 39.6 million times a month, compared with Gemma's 20.8 million and Llama's 7.5 million.

Why Qwen Won the Open-Source Race

Experts attribute Qwen's position to three factors: regular updates, coverage of all scales from sub-billion versions to Qwen3.8-Max (2.4 trillion parameters), and the Apache 2.0 license, which does not restrict modification or commercial use. In China, 59% of models with more than 20 billion parameters are released under Apache 2.0, 22% under MIT, and none carry commercial-use restrictions. American developers show a different picture: 29% under Apache/MIT, 41% under proprietary terms, and 30% without a specified license.

The report showed a shift in balance: almost every month this year, the largest open model from China surpassed all US releases in size. Its ceiling ranged from 754 billion to 2.78 trillion parameters, while competitors' ceilings did not exceed 130 billion in five of the seven months.

Alibaba's Cloud Play Turns Downloads Into Revenue

Alibaba is pursuing an indirect monetization strategy: offering Qwen for free while steering enterprises that want to use it toward its cloud services. The company distributes Qwen directly to enterprise customers in emerging markets such as Southeast Asia and Africa through its cloud platform — a distribution channel many competitors lack. This is a classic platform business model: capture the developer community through open source, then convert that into cloud infrastructure revenue.

US export controls on semiconductors and AI systems have failed to slow the momentum of Chinese competitors. While there was a temporary restriction on overseas access to Anthropic's latest model this summer, analysts suggest this accelerated adoption of Chinese models. US tech companies are responding — Meta and Nvidia have released new open AI models in recent weeks, with Nvidia strengthening its open model strategy tied to its GPU business.

Industry observers interpret Qwen's rise to No. 1 as a signal that China's AI industry has moved beyond catch-up and is now setting standards for the global open-source market. For investors, the download figures suggest Alibaba's cloud business — the primary monetization channel for Qwen — could see accelerating enterprise adoption, while Meta and Alphabet face mounting pressure to differentiate their open models.

This article is for informational purposes only and does not constitute investment advice.