In this video I show you how to get an AI coding assistant working inside VSCode without being forced into an online-only setup like Windsurf or GitHub Copilot. The whole idea is simple: keep your workflow in the editor you already use, but run the model locally so you’ve got more control over privacy, costs, and what’s actually happening on your machine. I walk through the tools I’m using and how they fit together so you can replicate the setup quickly. I use VSCode as the editor, Continue as the extension layer that gives you the chat/assist experience, and LM Studio to run a local LLM. The main takeaway is that you can get a “Copilot-like” experience while keeping things local—no account drama, no constant cloud dependency, and a setup you can tweak as you learn. If you’re doing Python, C++, Nim, Linux work, or just general dev stuff, this kind of local AI setup can be a really practical upgrade to your day-to-day coding.