VS Code AI Toolkit - AI Developer Tools Tool
Overview
AI Toolkit for Visual Studio Code is a Microsoft-maintained VS Code extension that centralizes model discovery, prompt engineering, agent building, local model runtime and basic deployment workflows inside the editor. The extension provides a Model Catalog (GitHub, OpenAI, Anthropic, Google, Foundry-hosted models), a Playground for interactive testing (including multi-modal attachments), an Agent/Prompt Builder with function-calling and MCP tool integration, plus tools for bulk runs, evaluation, conversion and fine-tuning. ([github.com](https://github.com/microsoft/vscode-ai-toolkit)) The extension targets developers, AI engineers and researchers who want to iterate quickly (local or cloud), compare models side-by-side, and build agents that can call external tools via MCP or be deployed to Microsoft Foundry. It supports running models locally (ONNX, Ollama) and cloud workflows (fine-tune locally or via Azure Container Apps), and includes profiling/tracing features for performance troubleshooting. The extension is published on the Visual Studio Marketplace and is actively updated with new model integrations and platform features. ([marketplace.visualstudio.com](https://marketplace.visualstudio.com/items?itemName=ms-windows-ai-studio.windows-ai-studio))
GitHub Statistics
- Stars: 1,870
- Forks: 197
- Contributors: 18
- License: MIT
- Last Updated: 2026-01-06T07:01:42Z
The GitHub repository shows steady project activity and an active Microsoft-maintained codebase: approximately 1.9k stars, ~197 forks, 18 contributors, and an issues tracker with ~97 open issues, indicating active community engagement and ongoing work. The repo includes a comprehensive README and changelog used to surface feature updates. These metrics point to a well-adopted open-source extension with an active issue backlog that the maintainers review. ([github.com](https://github.com/microsoft/vscode-ai-toolkit))
Installation
Install via npm:
code --install-extension ms-windows-ai-studio.windows-ai-studioext install ms-windows-ai-studio.windows-ai-studioIn VS Code: Extensions view → search for "AI Toolkit for Visual Studio Code" → Install (or use Quick Open Ctrl+P and paste the ext install command). Key Features
- Model Catalog: browse and compare models from GitHub, OpenAI, Anthropic, Google, Foundry, ONNX and Ollama.
- Playground: interactive chat-style testing with multi-modal attachments (images, audio) for model experimentation.
- Agent Builder: prompt-first agent authoring with function-calling, prompt chaining, variables and versioning.
- MCP tool integration: connect agents to Model Context Protocol (MCP) servers to call external tools and APIs.
- Bulk Run: run the same prompts across multiple models or test cases for comparative evaluation.
- Evaluation: compute F1, relevance, similarity and coherence metrics; add custom evaluators against ground truth.
- Model conversion & profiling: convert/quantize models for local deployment and profile CPU/GPU/NPU usage (Windows ML).
Community
AI Toolkit has a large install base on the Visual Studio Marketplace (Marketplace listing shows hundreds of thousands of installs) and a vibrant GitHub project with ~1.9k stars, ~197 forks and active issues and contributors. Documentation and Microsoft blogs provide tutorials and roadmap notes (for example Ollama/local model support and Foundry deployment), while the GitHub issues list shows practical user reports (installation, model download and platform-specific questions) that the maintainers and community address. Overall the project is actively maintained by Microsoft and used widely across developer and AI engineering communities. ([marketplace.visualstudio.com](https://marketplace.visualstudio.com/items?itemName=ms-windows-ai-studio.windows-ai-studio))
Key Information
- Category: Developer Tools
- Type: AI Developer Tools Tool