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.

AI Copilot for Vim/NeoVim

Unified management and routing for llama.cpp, MLX and vLLM models with web dashboard.

Build an Autonomous Web3 AI Trading Agent (BASE + Uniswap V4 example)

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

Reinforcement learning for text generation on MLX (Apple Silicon)

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

Practical guide to running LLMs locally on Apple Silicon, with a focus on architecture, memory scaling, and agent workflows.

Add MLX support to Pydantic AI through LM Studio or mlx-lm, run MLX compatible HF models on Apple silicon.

Federated Fine-Tuning of LLMs on Apple Silicon with Flower.ai and MLX-LM

Various LLM resources and experiments

在 macOS 上把本地模型包装成 OpenAI 风格 API（Chat + TTS），方便你的应用/SDK 直接接入。｜Wrap local models into an OpenAI-style API (Chat + TTS) on macOS, so any OpenAI-compatible client can connect.

A comprehensive toolkit for end-to-end continued pre-training, fine-tuning, monitoring, testing and publishing of language models with MLX-LM

Run large Mixture-of-Experts LLMs that exceed system RAM on Apple Silicon by loading only router-selected experts from SSD with MLX. Includes OpenAI/Anthropic-compatible serving for local agentic coding.

llmost - the most easy llm host

Swift app to summarize YouTube content using AI

Benchmark Gemma 4 E2B on Apple Silicon: MLX (mlx-lm) vs GGUF (llama-server), with TTFT, tokens/sec, and memory.

Train reasoning models on your Mac. GRPO training framework for Apple Silicon with curriculum learning.

Drop-in KV cache compression for MLX on Apple Silicon. Brings PolarQuant (Google, ICLR 2026) to mlx-lm with first-class Gemma 4 support: MatFormer, dual head_dim, hybrid sliding/global attention, cross-layer KV sharing. 3-bit → 4.8× smaller cache, 0.995 logit cosine @ 4-bit.

LLM model inference on Apple Silicon Mac using the Apple MLX Framework.