nativ

Blaizzy

Nativ 是一款原生 macOS 应用，可在 Apple Silicon 上本地运行 AI 模型，提供聊天、模型管理、性能分析以及与 OpenAI 和 Anthropic 兼容的 API 服务器。

Stars

890

7 天增长

暂无数据

Fork 数

开放 Issue

开源协议

MIT

为什么值得关注

它提供了一个精美的全功能 SwiftUI 工作区，集成了模型发现、高级推理控制，并直接支持 Codex、Claude Code 等流行编程工具。

适合谁使用

希望私密、本地运行 AI 的 Apple Silicon Mac 用户
需要兼容 OpenAI/Anthropic 的本地推理服务器的开发者
管理并试验 MLX 模型的 AI 爱好者
寻找本地后端的编程代理（Codex、Claude Code 等）用户

典型使用场景

使用本地模型进行私密聊天和视觉对话，支持持久化历史记录
运行兼容 OpenAI 和 Anthropic 的本地 API 服务器，供其他应用调用
从 Hugging Face 发现、下载和管理 MLX 模型
监控模型性能，包括 Token 用量、解码速度及请求分析

项目优势

原生 SwiftUI 体验，界面精美且支持菜单栏控制
内置 mlx-vlm 服务器，兼容 OpenAI 和 Anthropic API
性能仪表盘跟踪请求量、Token 指标及首字节时间
集成编程工具（Codex、Claude Code、Pi、Hermes、OpenCode）用于本地推理

使用前须知

需要 macOS 26+ 及 Apple Silicon，限制了兼容性
首次设置和模型下载需要网络连接
专用音频生成和图像生成模型尚不支持（即将推出）

README 快速开始

Nativ

Local AI, native to your Mac.

Chat, serve, monitor, and connect MLX models from one macOS app.

Nativ is a native macOS workspace for running AI models locally on Apple silicon. It bundles an mlx-vlm server, finds compatible models in your Hugging Face cache, and wraps the whole experience in a polished SwiftUI app.

Use Nativ as a private chat app, a model manager, a performance dashboard, or an OpenAI- and Anthropic-compatible local inference server for the tools you already use.

What Nativ can do

Feature	What you get
Local chat and vision	Streaming conversations, image attachments, reasoning output, response metrics, and persistent chat history.
Model library	Discover installed MLX models, browse compatible models on Hugging Face, download them, inspect capabilities, switch models, or remove old ones.
Performance analytics	Track request volume, token usage, time to first token, decode speed, model performance, and recent activity.
Local APIs	OpenAI-compatible chat, Responses, image, audio, and model endpoints, plus Anthropic Messages endpoints.
Coding-tool integrations	Configure and launch Codex, Claude Code, Pi, Hermes, and OpenCode against models served by Nativ.
Developer workspace	Inspect runtime details, copy endpoint URLs, search and filter live server logs, and monitor server health.
Menu bar controls	Start or stop the server, change the loaded model, check serving statistics, and open the main app without breaking focus.
Advanced inference controls	Tune sampling, thinking budgets, structured output, KV-cache quantization, prefix caching, and speculative decoding.

Inference runs on your Mac after a model has been downloaded. Model downloads and first-time build dependencies still require network access.

Coming soon

Support for dedicated audio-only and image-generation-only models is coming soon.

How it works

flowchart LR
    A["Nativ · SwiftUI app"] --> B["NativServerKit"]
    B --> C["Bundled mlx-vlm server"]
    C --> D["MLX runtime"]
    D --> E["Local models · Apple unified memory"]
    F["Apps and coding agents"] -->|"localhost API"| C

NativServerKit owns the embedded Python distribution and server lifecycle. The app adds model discovery, chat, a

项目描述

Local AI, native to your Mac. Chat, serve, monitor, and connect MLX models from one macOS app.