OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Infrastructure & Serving

Mlx Openai Server

Mlx Openai Server is a Python API server that runs local MLX models on Apple Silicon and exposes them through OpenAI-compatible endpoints. It supports language and multimodal inference, image generation and editing, embeddings, and Whisper audio transcription.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Mlx Openai Server — OpenViST project profile in AI Infrastructure & Serving.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you have an Apple Silicon Mac and want to run an MLX model locally behind an API interface familiar to OpenAI SDK clients.

Who it is for

  • Developers deploying MLX models locally on Apple Silicon.
  • Teams integrating local text, vision, image, embedding, or transcription models into software that uses OpenAI-compatible client patterns.

Documented uses

  • Serve a local language model through chat-completion or responses endpoints.
  • Send text and image inputs to a multimodal model.
  • Generate or edit images through dedicated endpoints.
  • Create embeddings or transcribe audio with a Whisper model.
Before you choose

Requirements & limits

The repository is public and identifies the project as MIT-licensed. The README requires macOS on Apple Silicon and Python 3.11 or later.

  • The host platform must be macOS on Apple Silicon.
  • Select the launch model type to match the workload; each type maps to a distinct backend and endpoint family.
  • The README documents local paths and Hugging Face repositories as model-path choices.
  • Whisper transcription also requires ffmpeg.
  • The featured Darwin 36B Opus configuration requires reasoning and tool-call parser options for those outputs to be parsed correctly.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install and launch a text-model server

Install the package and start a local language-model server for OpenAI-compatible clients.

Before you start

  • macOS on Apple Silicon
  • Python 3.11 or later
  • uv
  • An MLX model path or Hugging Face model repository

Steps

  1. Create and activate a Python 3.11 virtual environment.

  2. Install mlx-openai-server with uv pip.

  3. Run mlx-openai-server launch with --model-type lm and the chosen --model-path.

  4. Configure an OpenAI-compatible client to use http://localhost:8000/v1.

Expected outcome

A local text-model API endpoint that OpenAI-compatible clients can address.

Guide limitations

  • The README's example uses port 8000 and a non-empty placeholder API key.
  • Model-specific parser options may be needed for correct reasoning or tool-call parsing.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
cubist38/mlx-openai-server · Official homepage title · retained in its source language
How we review sourcesSuggest a correction