Data from LiveBench, an independent platform benchmarking large language models (LLMs), shows that Alibaba’s latest AI model, Qwen3, has surpassed competitor R1 to become the top open-source AI model globally. This milestone highlights the rapid development and increasing leadership of China in the field of artificial intelligence.

Alibaba’s cloud computing division released the Qwen3 family last week, comprising eight enhanced models ranging from 600 million to 235 billion parameters. In machine learning, parameters are the variables that help AI systems learn and generate accurate outputs based on data prompts. The new models have shown superior performance in various tests, including coding, mathematics, data analysis, and language instruction.

Prior to this breakthrough, DeepSeek’s R1 had held the top position on LiveBench’s rankings since its debut in January. The rise of Qwen3 reflects not only China’s technological advancements but also Alibaba’s growing influence in the global open-source AI community.

Open-source technologies have been a significant driver of China’s tech industry growth over the past decades. They allow developers worldwide to access, modify, and enhance AI models, fostering innovation and collaboration across borders.

China’s progress in AI is further exemplified by this recent achievement, indicating a strong upward trajectory in innovation and competitiveness on the global stage. Alibaba’s success in developing such a powerful open-source AI model underscores the country’s emerging dominance in artificial intelligence research and application.