Alibaba Cloud, also known as Aliyun – cloud computing company, a subsidiary of Alibaba Group has introduced Qwen 3.5. It is a new series of AI models developed for autonomous task execution. The update includes “visual agentic capabilities” designed to allow the system to interact with mobile and desktop applications to complete multi-step processes such as research or service bookings.
The model is natively multimodal, trained on combined visual and textual data to process images and video content up to two hours in duration. It supports over 200 languages, with expanded coverage for South Asian and African dialects. According to Alibaba’s internal benchmarking, Qwen 3.5-Plus performed at or above the levels of GPT-5.2 and Claude 4.5 in specific categories such as knowledge reasoning and instruction following.
In an effort to increase adoption, Alibaba has made an open-source version, Qwen 3.5-397B-A17B, available alongside its commercial offerings. The release occurs during a period of high competition in the Chinese AI sector, as firms like Alibaba, ByteDance, and DeepSeek compete for market share.
