What is opencode-lmstudio-warm?

opencode-lmstudio-warm is a plugin for opencode, an AI coding tool, that makes sure a local AI model running through LM Studio is fully loaded and ready before opencode sends it any request. Without this plugin, people running local models through LM Studio tend to hit three annoying problems: the very first request hangs while the model loads for the first time, requests fail outright with a no model loaded error if automatic loading is turned off, and models can get unloaded partway through a work session because LM Studio automatically frees up memory after a period of inactivity. The plugin fixes this by checking, before every request, whether the model is actually loaded, and if it is not, it loads the model first and only then lets the request through. It is careful to avoid loading the same model twice even if multiple requests come in around the same time. The project has been tested against specific versions of opencode and LM Studio on macOS with Apple Silicon. Setup involves running one command to register the plugin with opencode, either for a single project or globally on your machine, then pointing opencode's configuration at your local LM Studio server address. There is also a one-time settings change needed inside the LM Studio app itself, to turn off the automatic idle timeout that would otherwise unload models on its own. The plugin also offers a range of optional settings, like how long a loaded model should stay resident, how many requests it should handle at once, and how it should behave if a model fails to load, though it works out of the box with sensible defaults and no configuration required. This is a tool for developers who are already using opencode with local models through LM Studio rather than a cloud AI service, and it does not require any external dependencies beyond opencode and LM Studio themselves.