Build AI Agents with MCP Tool Use in Minutes with AI Toolkit for VSCode
We’re excited to announce Agent Builder, the newest evolution of what was formerly known as Prompt Builder, now reimagined and supercharged for intelligent app development. This powerful tool in AI Toolkit enables you to create, iterate, and optimize agents—from prompt engineering to tool integration—all in one seamless workflow. Whether you're designing simple chat interactions or complex task-performing agents with tool access, Agent Builder simplifies the journey from idea to integration. Why Agent Builder? Agent Builder is designed to empower developers and prompt engineers to: 🚀 Generate starter prompts with natural language 🔁 Iterate and refine prompts based on model responses 🧩 Break down tasks with prompt chaining and structured outputs 🧪 Test integrations with real-time runs and tool use such as MCP servers 💻 Generate production-ready code for rapid app development And a lot of features are coming soon, stay tuned for: 📝 Use variables in prompts �� Run agent with test cases to test your agent easily 📊 Evaluate the accuracy and performance of your agent with built-in or your custom metrics ☁️ Deploy your agent to cloud Build Smart Agents with Tool Use (MCP Servers) Agents can now connect to external tools through MCP (Model Control Protocol) servers, enabling them to perform real-world actions like querying a database, accessing APIs, or executing custom logic. Connect to an Existing MCP Server To use an existing MCP server in Agent Builder: In the Tools section, select + MCP Server. Choose a connection type: Command (stdio) – run a local command that implements the MCP protocol HTTP (server-sent events) – connect to a remote server implementing the MCP protocol If the MCP server supports multiple tools, select the specific tool you want to use. Enter your prompts and click Run to test the agent's interaction with the tool. This integration allows your agents to fetch live data or trigger custom backend services as part of the conversation flow. Build and Scaffold a New MCP Server Want to create your own tool? Agent Builder helps you scaffold a new MCP server project: In the Tools section, select + MCP Server. Choose MCP server project. Select your preferred programming language: Python or TypeScript. Pick a folder to create your server project. Name your project and click Create. Agent Builder generates a scaffolded implementation of the MCP protocol that you can extend. Use the built-in VS Code debugger: Press F5 or click Debug in Agent Builder Test with prompts like: System: You are a weather forecast professional that can tell weather information based on given location. User: What is the weather in Shanghai? Agent Builder will automatically connect to your running server and show the response, making it easy to test and refine the tool-agent interaction. AI Sparks from Prototype to Production with AI Toolkit Building AI-powered applications from scratch or infusing intelligence into existing systems? AI Sparks is your go-to webinar series for mastering the AI Toolkit (AITK) from foundational concepts to cutting-edge techniques. In this bi-weekly, hands-on series, we’ll cover: 🚀SLMs & Local Models – Test and deploy AI models and applications efficiently on your own terms locally, to edge devices or to the cloud 🔍 Embedding Models & RAG – Supercharge retrieval for smarter applications using existing data. 🎨 Multi-Modal AI – Work with images, text, and beyond. 🤖 Agentic Frameworks – Build autonomous, decision-making AI systems. Watch on Demand Share your feedback Get started with the latest version, share your feedback, and let us know how these new features help you in your AI development journey. As always, we’re here to listen, collaborate, and grow alongside our amazing user community. Thank you for being a part of this journey—let’s build the future of AI together! Join our Microsoft Azure AI Foundry Discord channel to continue the discussion 🚀Exploring Azure AI Model Inference: A Comprehensive Guide
Azure AI model inference provides access to a wide range of flagship models from leading providers such as AI21 Labs, Azure OpenAI, Cohere, Core42, DeepSeek, Meta, Microsoft, Mistral AI, and NTT Data https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models . These models can be consumed as APIs, allowing you to integrate advanced AI capabilities into your applications seamlessly. Model Families and Their Capabilities Azure AI Foundry categorises its models into several families, each offering unique capabilities: AI21 Labs: Known for the Jamba family models, which are production-grade large language models (LLMs) using AI21's hybrid Mamba-Transformer architecture. These models support chat completions, tool calling, and multiple languages including English, French, Spanish, Portuguese, German, Arabic, and Hebrew. https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Azure OpenAI: Offers diverse models designed for tasks such as reasoning, problem-solving, natural language understanding, and code generation. These models support text and image inputs, multiple languages, and tool calling https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Cohere: Provides models for embedding and command tasks, supporting multilingual capabilities and various response formats https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Core42: Features the Jais-30B-chat model, designed for chat completions https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models DeepSeek: Includes models like DeepSeek-V3 and DeepSeek-R1, focusing on advanced AI tasks https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Meta: Offers the Llama series models, which are instruction-tuned for various AI tasks https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Microsoft: Provides the Phi series models, supporting multimodal instructions and vision tasks https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Mistral AI: Features models like Ministral-3B and Mistral-large, designed for high-performance AI tasks https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models NTT Data: Offers the Tsuzumi-7b model, focusing on specific AI capabilities https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models Deployment and Integration Azure AI model inference supports global standard deployment, ensuring consistent throughput and performance. Models can be deployed in various configurations, including regional deployments and sovereign clouds such as Azure Government, Azure Germany, and Azure China https://learn.microsoft.com/azure/ai-foundry/model-inference/concepts/models To integrate these models into your applications, you can use the Azure AI model inference API, which supports multiple programming languages including Python, C#, JavaScript, and Java. This flexibility allows you to deploy models multiple times under different configurations, providing a robust and scalable solution for your AI needs https://learn.microsoft.com/en-us/azure/ai-foundry/model-inference/overview Conclusion Azure AI model inference in Azure AI Foundry offers a comprehensive solution for integrating advanced AI models into your applications. With a wide range of models from leading providers, flexible deployment options, and robust API support, Azure AI Foundry empowers you to leverage cutting-edge AI capabilities without the complexity of hosting and managing the infrastructure. Explore the Azure AI model catalog today and unlock the potential of AI for your business. Join the Conversation on Azure AI Foundry Discussions! Have ideas, questions, or insights about AI? Don't keep them to yourself! Share your thoughts, engage with experts, and connect with a community that’s shaping the future of artificial intelligence. 👉 Click here to join the discussion!
