mlx-vlm

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

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.

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.

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

Various LLM resources and experiments

MLX Porting Toolkit — an agent-guided, evidence-gated pipeline (scaffold → convert → parity → benchmark) plus a portable skill for porting PyTorch/Hugging Face models to Apple MLX.

Training-free decision engine for existing MLX models on Apple Silicon. No fine-tuning or extra decision weights. MLX-LM + experimental MLX-VLM action selection and semantic routing. Related: AnyJev.

Unlimited-OCR on Apple Silicon via MLX MXFP8 (mlx-vlm 0.6+). Includes config fix for native unlimited-ocr loading.

Local multimodal AI for Apple Silicon: chat with Gemma 4 via MLX, serve it over an OpenAI-compatible API, or run your own mlx-community model. On-device voice, documents, and agentic web research with citations.

Fast local LLM / VLM inference engine for Apple Silicon (MLX). OpenAI- and Anthropic-compatible, lossless speculative decoding (DFlash2 / MTP), hybrid-model prefix cache with RAM/SSD tiers, native Claude Code / Codex / opencode support. Qwen3.8-27B tuned first.

Run Pi coding agent isolated in a Docker Sandbox microVM with a local mlx-lm inference backend.

macOS menu bar app to start, stop and switch local LLM servers (mlx-lm, Ollama) — no terminal needed.

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.

Streamlit application for generating and editing images using Black Forest Labs FLUX.2 Klein on Apple Silicon with MLX.

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

GLM-OCR fine-tuned for Italian documents and quantized to 8 bits for MLX on Apple Silicon: Italian +3.7, English +7.1 points, 45% of the size