LocalAI QuickStart: Run OpenAI-Compatible LLMs Locally

LocalAI is a self-hosted inference server providing OpenAI API-compatible endpoints for running LLMs and multi-modal AI workloads (text, images, audio, embeddings) on local hardware. It supports multiple backends with CPU and GPU acceleration options, and is most easily deployed via Docker containers with various image flavors tailored to different hardware configurations including NVIDIA CUDA, AMD ROCm, and Intel oneAPI.