Alibaba Unveils Qwen 3.8 AI Model With 2.4 Trillion Parameters at Shanghai Summit

Alibaba has introduced its latest artificial intelligence model, Qwen 3.8, at the World AI Conference in Shanghai. The company claims Qwen 3.8 is among the most advanced AI models, second only to Anthropic's leading model. This announcement highlights the rapid progress of Chinese AI labs as they compete with American counterparts.

Key Highlights

• Alibaba introduced Qwen 3.8 AI model with 2.4 trillion parameters at the World AI Conference in Shanghai.
• Qwen 3.8 outperforms previous versions in coding, reasoning, and complex software development tasks.
• Moonshot AI's Kimi K3 model has 2.8 trillion parameters and claims higher benchmark performance.
• Performance claims from both Alibaba and Moonshot AI have not yet been independently verified.

Qwen 3.8 Model Features

Qwen 3.8 contains 2.4 trillion parameters, which measure the complexity of an AI model. Alibaba reports that Qwen 3.8 surpasses its previous version, Qwen 3.7 Max, in coding, reasoning, and complex tasks. These tasks include full-stack software development, data analysis, and office workflows.

The model is currently available for preview on Alibaba's Token Plan subscription service and its Qoder and QoderWork platforms. Alibaba plans to release Qwen 3.8 as an open-weight model soon. This will allow developers access to the model's learned parameters, though its training data and source code will remain proprietary.

Competition in China's AI Sector

Alibaba's launch comes shortly after Moonshot AI, a Beijing-based lab, introduced its own model, Kimi K3. Kimi K3 features 2.8 trillion parameters, making it larger than Qwen 3.8. Moonshot AI claims Kimi K3 delivers high performance across various benchmarks and outperforms Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol in software project development and long-horizon engineering tasks.

However, these performance claims from both Alibaba and Moonshot AI rely on self-reported data. Independent benchmark organizations have not yet verified these results. Alibaba also states that Qwen 3.8 is stronger than Kimi K3, but this has not been independently confirmed.

Broader Trends and Political Context

The recent launches indicate that Chinese AI labs have entered the multi-trillion parameter era. They are now benchmarking their models against leading global AI companies, including Anthropic, OpenAI, and Google. This trend reflects the growing competitiveness of China's AI industry.

The launches also carry political significance. Chinese President Xi Jinping attended the World AI Conference in person this year. He emphasized China's commitment to open source, openness, collaboration, and sharing in the AI sector. President Xi stated that this approach aims to ensure all sectors and businesses can benefit from AI advancements.