Prime Highlights :
- Alibaba launched Qwen3.8-Max, its largest AI model, with support for text, image and video processing.
- The model uses a mixture-of-experts design that improves efficiency by activating only a small portion of its total parameters for each task.
Key Facts :
- Alibaba is a Chinese technology company with businesses in e-commerce, cloud computing and artificial intelligence.
- Qwen3.8-Max has 2.4 trillion parameters and can process up to one million tokens in a single request.
Background :
Alibaba has rolled out Qwen3.8-Max, its biggest, most advanced AI model, as competition keeps heating up between Chinese tech companies all around. The new model has 2.4 trillion parameters, placing it close to domestic rival Moonshot AI’s Kimi K3, which has 2.8 trillion parameters. While parameter count does not directly determine performance, it remains an important measure of a model’s scale and computing capability.
Qwen3.8-Max supports text, image and video processing and can handle up to one million tokens in a single request. This allows users to analyse large software codebases, lengthy legal documents and hundreds of pages of content without splitting the input.
Alibaba said the model uses a mixture-of-experts architecture that activates only 95 billion parameters for each task instead of running the full model. The company said this design lowers computing costs and reduces response time while maintaining strong performance. It also stated that the model completed a software engineering project in 16 days.
The company introduced Qwen3.8-Max on the Arena.AI model comparison platform, where it became the highest-ranked Chinese text model. It still trails Anthropic’s Claude Fable 5 and several Opus models in overall text rankings. In visual AI rankings, however, the model secured the second position globally, behind only a Claude Fable 5 variant.
Alibaba plans to release Qwen3.8-Max through its Alibaba Cloud Model Studio platform next week, expanding access for developers building AI applications.