Description
Local LLM applications can expose an OpenAI-compatible API without relying on a cloud model provider. This CPU-only server is useful for developers testing tools against local models on machines without GPU acceleration. Prompts and generated text stay local only if clients and model files are configured that way.