ZedFrom the builder
Why and How to Run Local Models in Zed
Zed lays out the case for using local models in its editor and how to set them up, noting that local model usage in Zed's agent has tripled over the past 10 weeks. The post argues local models are fully private, cheaper over time, freely configurable and immune to provider changes, while conceding that models runnable on consumer hardware still trail those from top labs. The author has had most success with LM Studio, with Ollama and llama.cpp as other popular runtimes.
Models you will be able to run locally are not as capable as what you can get from the top AI labs.
Saved . A shortened excerpt from the official blog; the original publisher retains authorship.