A high-performance API server that provides OpenAI-compatible endpoints for MLX models. Developed using Python and powered by the FastAPI framework, it provides an efficient, scalable, and user-friendly solution for running MLX-based vision and language models locally with an OpenAI-compatible interface.

macMLX brings local LLM inference to Apple Silicon with a first-class native macOS experience. No cloud, no telemetry, no Electron — just your Mac running models at full speed.

Experimental: MLX model provider for Strands Agents - Build, train, and deploy AI agents on Apple Silicon.

Real-time webcam demo with SmolVLM(mlx-community/SmolVLM-Instruct-4bit) and MLX-VLM

An all-in-one LLMs chat Web UI based on the MLX framework, designed for Apple Silicon.

Defensive layer for mlx / mlx-lm / mlx-vlm on Apple Silicon. Prevents IOGPUFamily kernel panics, SIGABRT, and Mac reboots from MLX inference. Includes a community-curated registry of known-panic models (KNOWN_PANIC_MODELS) with hardware contexts, root-cause hypotheses, and verified workarounds.

Local AI for Apple Silicon. Chat with open LLMs (Qwen, Gemma, DeepSeek, Mistral, Llama…) fully offline on your Mac via Apple MLX — private by design, zero telemetry.

Various LLM resources and experiments

llmost - the most easy llm host

RETIX is a local-first vision MLX based CLI for agents that need to inspect screenshots, extract visible text, and verify visual claims with deterministic output.

A lightweight, config-driven framework for unified vision-language model inference across local and cloud backends.

Generate and edit images with the Black Forest Labs FLUX.2 Klein 4B model on Apple Silicon with MLX.

MLX inference service compatible with OpenAI API, built on MLX-LM and MLX-VLM.基于MLX-LM和MLX-VLM构建的OpenAI API兼容的MLX推理服务.

🤖 Run Strands Agents on Apple Silicon with ease—perform inference, fine-tune models, and leverage vision capabilities using Python and LoRA training.

OpenAI-compatible Qwen3-Omni server on Apple Silicon (MLX / mlx-vlm). Text + image + audio + video in, text / tool calls out. Single-flight queue, resident model, sync-only.

Repeatable Apple Silicon local LLM benchmark scripts, reports, and docs

OpenAI-compatible multi-model Qwen server for Apple Silicon (MLX / mlx-vlm): Qwen3-Omni + Qwen3.6-27B in one process, memory-budgeted resident cache, text/image/audio/video in, text/tool-calls out.

Extract structured information from documents, convert images to Markdown, and generate templates with the NuMind NuExtract3 model on Apple Silicon with MLX.

Analyze medical text, 2D images (e.g. chest X-ray), 3D CT volumes, and whole-slide pathology images with the Google MedGemma 1.5 model on Apple Silicon with MLX.

This repository provides a python script for running Vision Language Models via mlx-vlm