In August 2026, the Chinese tech giant Alibaba Group Holding reported unprecedented success for its artificial intelligence ecosystem. The Qwen family of open models reached a milestone of 3 billion downloads worldwide in just the last six months. This figure allowed Alibaba not only to take a leading position but to confidently overtake Western competitors like Google and Meta in the segment of open-source language models.
New Leader in Open AI: Figures and Facts
Statistics published by Alibaba demonstrate the scale of the phenomenon. The Qwen family currently includes more than 460 different neural networks, and the entire ecosystem, including derivative algorithms, counts over 300,000 versions. For comparison, according to Hugging Face data, during the same period (the first half of 2026), Google models were downloaded 418 million times, and Meta neural networks — 227 million times. A gap of more than 10 times in favor of the Chinese company indicates a paradigm shift in the open AI market.
The success of Qwen is explained not only by the number of models but also by their accessibility. Open models allow developers to freely download, customize, and use them as a basis for their own products. This creates a "snowball effect": the more developers use the base model, the more improved versions appear, which in turn attracts new users.
Expansion Strategy and Bypassing Restrictions
The growth in popularity of Qwen is taking place against the backdrop of a fierce US-China technology race. Despite US export controls and restrictions on the supply of advanced AI accelerators to China, Chinese developers demonstrate high adaptability. Alibaba's strategy is built on promoting powerful yet relatively cheap and easily adaptable models.
The company is actively developing its presence outside of China, using its cloud platform for corporate clients. Special attention is paid to the markets of Southeast Asia and Africa, where Alibaba achieves coverage unavailable to many Western competitors. Experts note that Qwen has become part of the standard workflow for developers worldwide, deciding which models to fine-tune and deploy.
Competition with Closed Models
In addition to the battle for open source, Chinese developers, including Alibaba, Moonshot AI Inc., and DeepSeek, are striving to close the gap with the closed neural networks of American companies such as OpenAI and Anthropic. Data on Qwen model downloads, according to Hugging Face analysts, makes Alibaba's algorithms "one of the main elements of the foundation of the open AI ecosystem." This confirms that the Chinese strategy of creating a self-developing ecosystem is gaining momentum and becoming a serious challenge for traditional technology leaders.
