Alibaba Group has launched Qwen3-Coder, an open-source artificial intelligence model tailored for software development tasks, as global competition intensifies in the AI coding tool space.

In a statement, Alibaba said the new model is designed for code generation and to manage programming workflows. It is also capable of performing what the company terms “agentic coding” tasks, allowing the AI to independently tackle coding challenges without human intervention.

The headline model, Qwen3-Coder-480B-A35B-Instruct, operates with 480 billion total parameters, of which 35 billion are actively used during execution. It supports up to 256,000 tokens of context, with scalability up to one million tokens—enabling it to manage complex programming tasks and large codebases.

According to Alibaba, the model can interact with code repositories and manage function calls in various programming environments. In benchmark tests, Qwen3-Coder has shown improved performance relative to some China-based models and has matched the output quality of specific US-developed AI systems.

The release comes as major technology firms continue to introduce advanced AI models for use in software development, expanding the role of artificial intelligence in programming and automation.

The company has not specified commercial plans for Qwen3-Coder but confirmed it will be made available as open-source.