The AI world is constantly shifting, with Anthropic looking to pull Claude Code from Pro plans and GitHub Copilot closing subscriptions to new users. Where is all this heading?
I bought my first MacBook a few months ago, and the first thing I wanted to figure out was how well it could run local models, how they stack up against the big commercial ones, and how to get the most out of them.
In the post below I walk you through how I set up LM Studio to run a model locally and hook it up to Visual Studio Code using the Continue plugin. I tested models with 30B parameters, which are nowhere near the level of something like Opus or Sonnet, but honestly the potential surprised me.
These days I’m working on a pretty interesting project: extracting business logic from a legacy application using a Claude Code skill, then reshaping and documenting it across several markdown files. That output is then picked up by another skill (on an Angular project) that handles creating the base components, services and stores, applying the logic extracted and documented in the first artifacts.
The Angular project skill relies on a set of templates and guardrails, and using less capable AI models could actually make a real difference here, helping us save a lot of tokens.
I’ll explain well this project in a follow-up article, stay tuned! In the meantime, if you want to run a local model with LM Studio, check out my article!
🔗 <a href="https://www.thinkasadev.com/en/local-ai-with-vs-code-and-continue-2/" target="_blank" rel="noopener nofollow">https://www.thinkasadev.com/en/local-ai-with-vs-code-and-continue-2/</a>

Giorgio

A setup guide for running local AI models using LM Studio on a MacBook and connecting them to Visual Studio Code via the Continue plugin. The author tested 30B parameter models and found them surprisingly capable. The post also briefly describes a real-world use case involving Claude Code for extracting business logic from a legacy app and generating Angular components, where local models could help reduce token costs.

Local AI: LM Studio + VS Code setup guide