Qwen3-Coder

You can install the Continue.AI extension directly from the Visual Studio Code marketplace. Once installed, you configure the extension's config file to route requests to the Qwen3-Coder model via an API provider or a local inference engine. It is specifically optimized for code completion tasks with its roughly 12 billion parameters.

As an open-source large language model developed by Alibaba Group, Qwen3-Coder performs exceptionally well across popular languages like Python, JavaScript, and Java. The model has been trained on vast repositories of code, allowing it to handle complex logic across dozens of languages. Its 12B parameter architecture makes it highly capable of understanding varied coding contexts.

Yes, because it is an open-source model, you can run it locally using inference tools like Ollama or LM Studio. You will need a machine with sufficient RAM or VRAM to handle the 12B parameters efficiently. Once running locally, you can point your Continue.AI extension to the local server address for completely private code generation.

Alibaba typically releases its Qwen series models under open-source licenses like Apache 2.0, which generally permit commercial use. However, you should always verify the specific terms of use associated with the Qwen3-Coder release on their official repository. If you access it through a paid API provider via Continue, standard token rates will apply instead.