Microsoft Foundry Local -- First Look and Walk Through

This video is a first look at Microsoft Foundry Local –a local inference engine introduced at Microsoft Ignite 2025. If you build AI apps and want to run models locally on Windows, macOS, or Windows Server 2025, this walkthrough gives you a complete first look at how Foundry Local works and what it enables. In this video, we’ll walk through: 🔍 What Microsoft Foundry Local Is A new runtime for hosting LLMs directly on your workstation or server, optimized for developers building secure and high-performance AI applications. ⚙️ Foundry Local Architecture Overview How the local model runtime, the OpenAI-compatible API layer, and the developer SDK fit together. 💻 Installing Foundry Local Step-by-step setup on a local Windows 11 machine, including model configuration and environment prep. 🚀 Testing Local Inference Testing inference and comparing performance with qwen2.5-0.5b and GPT-OSS-20B on an NVIDIA RTX 4090. 🧑‍💻 Building Local Apps with VS Code How to create applications that call your local models using the OpenAI-compatible SDK — perfect for rapid prototyping or offline development. 🎯 Who Is This For? AI developers, architects, and anyone evaluating: • On-prem or air-gapped AI workloads • Scenarios where data must remain local • Local experimentation before deploying to servers or cloud GPUs • A seamless path from prototype → production using OpenAI-compatible interfaces If you want a practical, developer-focused view of what Foundry Local can do — and how it fits into your workflow — this video is a great place to start. Getting Started with Foundry Local Link from Video: https://learn.microsoft.com/en-us/azure/ai-foundry/foundry-local/get-started Chapters 0:00 Introduction 1:22 What is Foundry Local? 2:27 Comparing to Alternatives 2:57 Installation 4:00 Initial test Qwen2 5:30 Load gpt-oss-20b 6:26 Test gpt-oss-20b 7:15 Create Python App