AI Toolkit for Visual Studio Code
AI Toolkit for Visual Studio Code is an extension to help developers and AI engineers to easily build AI apps through developing and testing with generative AI models locally or in the cloud. AI Toolkit supports most genAI models on the market.
AI engineers can use AI Toolkit to discover and try popular AI models easily with playground that has attachment support, run multiple prompts in batch mode, evaluate the prompts in a dataset to AI models for the popular evaluators, and fine-tune/deploy AI models.
Key features
- Model catalog with rich generative AI models sources (GitHub, ONNX, OpenAI, Anthropic, Google, ...)
- Bring Your Own Models from remotely hosted model, or Ollama models that are running locally
- Playground for model inference or testing via chat
- Attachment support for multi-modal language models
- Batch run prompts for selected AI models
- Evaluate an AI model with a dataset for supported popular evaluators like F1 score, relevance, similarity, coherence, and more
Who is AI Toolkit for?
Any developer who wants to explore, test, evaluate, and fine-tune generative AI models when building AI apps.
Install and setup
You can install the AI Toolkit from the Extensions view in VS Code:
You can switch the installation between the formal released version for stable features and pre-released version for early access of new features. Check What's New during the installation for detailed feature list of each version.
Get started with AI Toolkit
The AI Toolkit has a getting started walkthrough that you can use to learn the basics of the AI Toolkit. The walkthrough takes you through the playground, where you can use chat to interact with AI models.
-
Select the AI Toolkit view in the Activity Bar
-
In the Help and Feedback section, select Get Started to open the walkthrough
Next steps
- Get more information about adding generative AI models in AI Toolkit
- Use the model playground to interact with models