OpenCode now allows users to configure local LLMs, expanding its capabilities beyond cloud-based providers. This update lets developers leverage tools like vLLM, LM Studio, and Ollama, offering greater flexibility and control over their AI workflows. The ability to integrate local endpoints means users can experiment more freely and potentially reduce costs.

To set up a local LLM, users need to register the provider in OpenCode's authentication file and define the provider itself in the configuration file. The process involves specifying the API endpoint and model details. This approach, outlined in the guide, supports any OpenAI-compatible endpoint, making integration straightforward. It's a win for developers seeking local AI solutions.

This new functionality is particularly relevant as the demand for local AI processing grows. Running LLMs locally can enhance privacy, reduce latency, and enable offline use. The guide provides step-by-step instructions for wiring up a local vLLM server, but the same principles apply to other compatible endpoints.

Next steps involve restarting OpenCode and selecting the custom provider and model. The guide’s focus on OpenAI-compatible APIs streamlines the process, allowing for easy integration of diverse LLM solutions. This means developers have more freedom to choose the best models for their specific needs, whether for development or deployment.