en.Wedoany.com Reported - Dimension News, on the evening of August 14, Alibaba's Qwen large model officially open-sourced the Qwen3.8 series models, allowing all developers, research institutions, and enterprises to freely download, deploy, and use them. The core model released this time, Qwen3.8-27B, is a natively multimodal dense model with only 27 billion parameters, yet its overall performance surpasses Qwen3.7-Plus, delivering outstanding results in real-world programming and office scenarios.

The new model offers fast response speed, high output quality, and easy deployment, running smoothly on consumer-grade GPUs after quantization. Qwen3.8-27B is open-sourced under the permissive Apache 2.0 license and is expected to become a mainstream model for local deployment. Previously, the flagship model Qwen3.8-Max also had its model weights open-sourced.

Public data shows that Alibaba has cumulatively open-sourced over 460 models to the global AI community, with the Qwen series models surpassing 3 billion total downloads worldwide and over 300,000 derivative models, firmly ranking as the world's leading open-source model family.