Alibaba has launched Qwen3.6-Plus, the newest version of its flagship AI model, with major upgrades in coding and visual understanding.
The company said the new model is built for “agentic AI,” meaning it can go beyond simple responses and independently complete complex tasks, such as writing and refining software or analyzing real-world images.
Qwen3.6-Plus will be added to Alibaba’s platforms, including Wukong, an enterprise AI system that automates business tasks, and the Qwen App, its main AI application.
One of the model’s key features is its ability to perceive, think, and act in one workflow, helping developers turn ideas into finished products more efficiently. It can plan, test, and improve code on its own, acting like a full partner in software development.
The model also supports large-scale projects, with a context window of up to 1 million tokens, allowing it to handle complex coding tasks across entire repositories.
In addition to coding, Qwen3.6-Plus has improved multimodal capabilities, meaning it can understand and combine text, images, and video. For example, it can read documents, analyze visuals, and even turn design sketches or screenshots into working website code.
Alibaba said the model is designed for real-world use, offering stable and accurate performance for tasks like retail analysis and automated inspections.
Developers can access Qwen3.6-Plus through Model Studio, Alibaba Cloud’s AI platform, and use it with tools like OpenClaw, Claude Code, and Cline. The company also plans to release smaller open-source versions for developers.
