本地 AI 音乐生成
工作站
Local AI Music
Generation Studio

基于 ACE-Step 1.5 引擎的 macOS / Windows 桌面客户端。全本地推理,数据不出本机,一键生成歌曲并支持音轨编辑,为你的 MV 创作提供完整的音乐流水线。A macOS / Windows desktop client powered by the ACE-Step 1.5 engine. Fully local inference — your data never leaves the machine. Generate songs with one click, fine-tune tracks, and build a complete music pipeline for your MV projects.

下载 macOS / Windows 客户端Download macOS / Windows Client 查看源码View Source
macOS · Apple Silicon Windows 中文界面Chinese UI 本地推理Local Inference 音轨编辑Track Editing 自动拉起引擎Auto-launch Engine
ACE-Step Desktop 中文界面截图
为什么选择 ACE-Step DesktopWhy ACE-Step Desktop

🎵 全本地推理Fully Local Inference

ACE-Step 1.5 引擎在本地运行,模型权重与生成音频均不离开你的电脑,隐私安全,离线可用。The ACE-Step 1.5 engine runs entirely on your machine — model weights and generated audio never leave your computer. Private, secure, and works offline.

🌐 原生中文界面Native Chinese UI

内置 i18n 国际化支持,一键切换简体中文,无需修改代码。Built-in i18n support lets you switch to Simplified Chinese with one click — no code changes needed.

自动拉起服务One-Click Auto Launch

启动应用自动拉起推理引擎与后端服务,无需手动敲命令,打开即用。Launching the app automatically starts the inference engine and backend services. No command-line knowledge required — just open and go.

🎛️ 音轨编辑Track Editing

内置音频编辑器,支持人声分离、多轨处理,生成后直接精细打磨。A built-in audio editor with vocal separation and multi-track processing, so you can polish tracks right after generation.

🎬 面向 MV 工作流MV-First Workflow

生成歌曲后配合生图与视频合成工具,构建端到端的音乐 MV 生产管线。Pair generated songs with image and video tools to build an end-to-end music video production pipeline.

🖥️ 桌面原生体验Native Desktop Experience

Tauri 2 封装,轻量高效,独立窗口运行,不依赖浏览器标签页。Wrapped with Tauri 2 for a lightweight, efficient standalone window — no browser tabs required.

快速开始Quick Start
1

下载安装:从 GitHub Releases 下载对应安装包:macOS(Apple Silicon)或 Windows(.exe 安装程序)。Download & install: Grab the right package from GitHub Releases: macOS (Apple Silicon) or Windows (.exe installer).

2

首次启动:应用自动拉起 ACE-Step 1.5 推理引擎与 Node 后端,等待就绪后进入主界面。First launch: The app automatically starts the ACE-Step 1.5 inference engine and Node backend, then opens the main UI once ready.

3

创作歌曲:填写歌词与风格,选择参数,点击生成,等待本地推理完成。Create a song: Enter lyrics and style, pick your parameters, hit generate, and wait for local inference to finish.

4

编辑导出:在音轨编辑器中打磨音频,导出成品用于 MV 制作。Edit & export: Polish the audio in the track editor and export the final result for your MV.