Blaizzy · 2026-07-20 · notable
Nativ v0.0.1 — local AI Mac app that discovers and monitors MLX models
Nativ is a new open-source macOS app from mlx-vlm creator Prince Canuma that bundles an mlx-vlm server with a SwiftUI UI to chat with local MLX models, expose an OpenAI-compatible endpoint, and watch tokens/sec live.
A free, open-source Mac app that turns any MLX model on your disk into a local chat + OpenAI-style API.
What is it?
Nativ is a native macOS 26 app by Prince Canuma (Blaizzy) that discovers MLX models already on your Mac, lets you chat with them from a SwiftUI window, and exposes an OpenAI-compatible server so other tools can call the same model. Version 0.0.1 shipped on 2026-07-20 as a signed DMG.
How does it work?
Nativ bundles the mlx-vlm inference server and drives it from Swift. It scans local paths for MLX weights, lets you browse and download more from Hugging Face inside the app, and streams responses through a chat UI. A monitoring panel tracks request volume, token usage, time-to-first-token, decode speed, and per-model performance in real time.
Why does it matter?
Nativ gives Apple Silicon users a self-contained, free alternative to LM Studio or Ollama built directly on the MLX stack that Apple ships. Because the same process both serves inference and shows analytics, it removes the usual shell-plus-CLI dance for hobbyists who just want a chat window on top of an M-series GPU.
Who is it for?
Apple Silicon users who want a local chat UI on top of MLX-native models.
Try it
Download the DMG from https://github.com/Blaizzy/nativ/releases/latest