Alibaba's Qwen 3.8-27B tops Hugging Face with 1M downloads in 2 days

Alibaba's Qwen 3.8-27B has become the fastest-rising open-source AI model on Hugging Face, capturing the top spot on the global trend leaderboard and racking up more than 1 million downloads within 48 hours of its release. This 27B-parameter model is the latest addition to the Qwen family, which has grown into a sprawling ecosystem of 460+ open-source models and 300,000 derivative fine-tunes. The milestone underscores Alibaba's aggressive push to dominate the open-source AI landscape, a space traditionally led by Meta's Llama series and Google's Gemma models.

The Qwen family's total downloads have now surpassed 3 billion, a figure that eclipses Meta and Google's open-source offerings. Developers are flocking to Qwen for its strong multilingual performance, competitive benchmark results, and permissive licensing, which allows commercial use and customization. The rapid adoption of Qwen 3.8-27B signals that Alibaba is not just a participant in the global AI race but a foundational player, providing the building blocks for thousands of startups, enterprises, and researchers. With derivative models proliferating across industries—from coding assistants to domain-specific chatbots—Qwen is increasingly becoming the default choice for teams that want powerful, open, and cost-effective AI without relying on Western tech giants. This shift could reshape the balance of power in the AI ecosystem, particularly as demand for sovereign and regionally tailored models grows.