News Summary

Alibaba, a Chinese tech giant, has released Qwen 3, a new family of AI models that the company claims matches and in some cases outperforms the leading models from Google and OpenAI. The models, ranging from 0.6 billion to 235 billion parameters, are available under an open license on platforms like Hugging Face and GitHub. These ‘hybrid’ models are designed to handle both complex and simple tasks efficiently, with some utilizing a mixture of experts (MoE) architecture for better computational efficiency. Qwen 3 supports 119 languages and was trained on a diverse dataset of nearly 36 trillion tokens. Despite not being publicly available yet, the largest model, Qwen-3-235B-A22B, has shown competitive results in benchmarks, surpassing models like OpenAI’s o3-mini in several tests. This launch highlights China’s increasing role in the global AI landscape and adds pressure on American companies to maintain their technological superiority.