跳到正文

正在自动登录…

← 返回项目列表

MagoStudio/MagoStudio

视频剪辑视频调色模型本地部署

Mago Studio 是一款本地优先的桌面 AI 视频编辑器,支持脚本转视频、素材重风格化、帧精确编辑等功能。适合希望在本地 GPU 上运行 AI 视频编辑,或使用云模型进行渲染的用户。

AI 中文解读 · 根据仓库描述与 README 生成 · 2026/11/3 08:06:44 · 查看原文

值得看TypeScript应用与服务
累计 Stars
53
当前热度
1 天内 · 小火
仓库创建时间
打开 GitHub ↗

选用指南

AI 根据 README 整理 · 2026/11/3 08:06:44

适用场景

适合需要本地优先、AI 驱动的视频编辑工作流,包括脚本转视频、素材重风格化、帧精确编辑等场景。

查看 README 依据
The local-first desktop edition of the [Mago](https://mago.studio) AI video editor. Your projects and your videos stay on your machine. Renders run on your own GPU through ComfyUI, or in Mago Cloud when you sign in. A Mago Cloud render uploads only what that render needs: the Shot, its source video, reference images and masks. The rest of the Project stays local. See [docs-public/architecture.md](docs-public/architecture.md#3-local--cloud-wiring). ## Contents - [Features](#features) - [Screenshots](#screenshots) - [Made with Mago](#made-with-mago) - [Install](#install)
- [Run from source](#run-from-source) - [Documentation](#documentation) - [Security](#security) - [License](#license) ## Features - **Mago Director.** Script to video: script, cast, shot list and storyboards, then consistent scenes where each Shot picks up from the previous one. - **Mago Transform.** Footage in, restyled footage out, frame-accurate. Work on specific frame ranges, edit only part of a video with masks, relight, replace characters. - **More than clips.** Versioning, real review tools and a frame-accurate pipeline.
使用方式

通过安装程序使用,可选择登录 Mago 账户使用云模型,或跳过登录在本地 GPU 上离线工作。

查看 README 依据
</table> ## Install Mago Studio runs on Windows, macOS and Linux. You don't need a GPU to use the app or render in Mago Cloud. Local rendering is optional (step 3). 1. **Download** the installer for your OS from the [latest release](https://github.com/MagoStudio/MagoStudio/releases/latest). Windows builds are signed; macOS and Linux builds are unsigned. 2. **Open Mago Studio and choose how to work.** - **Sign in with your Mago account** to use Mago Cloud models alongside local ones. - **Press Skip** to work offline on your own GPU, with no account. The setup wizard
部署要求

Windows 上本地渲染需要 NVIDIA GPU、8 GB VRAM 和约 63 GB 可用磁盘空间;macOS 和 Linux 需要连接自己的 ComfyUI。

查看 README 依据
opens for you. You can sign in later. 3. Run **Set up local rendering**. On Windows the setup wizard installs ComfyUI, node packs and models for you (NVIDIA GPU, 8 GB VRAM minimum, about 63 GB free disk). On macOS and Linux, point Mago Studio at your own ComfyUI. > [!TIP] > **We recommend the portable ComfyUI the wizard installs.** It comes with the right > ComfyUI version and node packs already set up. Connecting an existing ComfyUI install > can cause problems with mismatched nodes or Python versions. ### What the setup does
已知限制

本地渲染需要特定硬件和磁盘空间,且连接现有 ComfyUI 可能因节点或 Python 版本不匹配而出现问题。

查看 README 依据
opens for you. You can sign in later. 3. Run **Set up local rendering**. On Windows the setup wizard installs ComfyUI, node packs and models for you (NVIDIA GPU, 8 GB VRAM minimum, about 63 GB free disk). On macOS and Linux, point Mago Studio at your own ComfyUI. > [!TIP] > **We recommend the portable ComfyUI the wizard installs.** It comes with the right > ComfyUI version and node packs already set up. Connecting an existing ComfyUI install > can cause problems with mismatched nodes or Python versions. ### What the setup does

README

正在加载 README…

收录信息

仓库创建时间
2026/11/3 00:03:10
首次达标收录
2026/11/3 08:01:18
统计更新时间
2026/11/3 09:01:22

按创建时长与累计 Stars 判断等级。首次达标收录后,即使未达当前门槛仍保留;累计 Stars 和仓库状态会定期更新。确认只有 README 及辅助文件的仓库会停止公开。

模型评估

值得看加权总分 4.0 / 5

新颖度
3.6
实用价值
4.3
成熟度
4.7
实质含量
3.5

由 Jev 模型根据仓库描述与 README 评估,2026/11/3 08:01:18 生成。评分仅供参考,不是代码、安全或许可证审查;热度等级另按创建时长与累计 Stars 计算。

仓库信息

最近推送
2026/11/3 02:45:38
许可证标识
AGPL-3.0
Fork 数
9
当前状态
未归档

收录记录

按创建时长与累计 Stars 首次收录:小火
首次发现项目