AI Toolkit for Visual Studio Code
The AI Toolkit for Visual Studio Code is a game-changer for developers and AI engineers looking to simplify the creation of AI applications. This powerful extension streamlines the entire process, from development and testing to deployment, by seamlessly integrating with generative AI models both locally and in the cloud. Supporting a wide range of commercially available generative AI models, the AI Toolkit empowers you to build cutting-edge AI solutions directly within your familiar VS Code environment.
Ready to get started? Simply search the Visual Studio Code extension marketplace for “AI Toolkit” and install it.

For instructions on installing Ollama and Deepseek-r1 locally on your machine, please refer to this guide.
Leveraging Your Local Ollama Installation
This guide will walk you through connecting the AI Toolkit to your existing local Ollama setup, allowing you to leverage models like Deepseek-r1 without leaving VS Code.
- Open the Command Palette: Use the keyboard shortcut
Cmd+Shift+P
(orCtrl+Shift+P
on Windows) to open the command palette. - Invoke AI Toolkit: Type “AI Toolkit” in the command palette.
- Add a Model for Remote Inference: Select “AI Toolkit: Add Model for remote inference” from the list of options.
- Choose Ollama Model: Select “Add an Ollama model.”
- Specify Ollama Endpoint: The input field will likely be pre-filled with
http://localhost:11434/
. If your Ollama instance is running on a different port, adjust this value accordingly. - Select Your Model: The AI Toolkit will now display a list of models currently installed in your local Ollama environment. Choose “deepseek-r1” (or any other model you wish to use) and click “OK.”
- Model Loading: The AI Toolkit will now load the selected model. This process may take a few seconds. Once complete, your VS Code environment will be ready to use the locally available deepseek-r1 (or your chosen model).
