mlx-openai-server
An OpenAI-compatible API server that runs MLX text, multimodal, image and Whisper models locally on Apple Silicon, exposed through the familiar OpenAI SDK endpoints.
mlx-openai-server serves MLX-based language, vision, image-generation, embedding and speech-recognition models on macOS with Apple Silicon behind an OpenAI-compatible HTTP API, so existing OpenAI SDK client code can point at a local server instead of a cloud endpoint. It requires Python 3.11 or newer and installs with pip or uv.
Supported model types include plain language models, multimodal vision-language models, image generation and editing through mflux, embeddings, and Whisper transcription, each launched with a model-specific configuration and optional reasoning and tool-call parsers.