header-langage
简体中文
繁體中文
English
Tiếng Việt
한국어
日本語
ภาษาไทย
Türkçe
Scan to Download the APP

Under the umbrella of Meta, Manus releases a desktop AI agent that steps out of the cloud to directly control the user's local computer.

According to 1M AI News monitoring, Manus, the general AI agent platform acquired by Meta at the end of last year, released the Manus Desktop desktop application yesterday, and the core feature "My Computer" brought Manus from the cloud sandbox to the user's local computer for the first time. It is now available to macOS and Windows users.

Since its launch, Manus has been running entirely in a cloud-based isolated sandbox with capabilities such as networking, command-line, file system, and browser, but it could not access user's local files and applications. My Computer operates on the local computer by executing command-line instructions on the user's terminal, allowing it to read, analyze, and edit local files, as well as launch and control local applications. The official blog post demonstrates scenarios such as having Manus automatically identify and categorize thousands of messy photos, and building a real-time meeting translation app from scratch on a Mac using Swift, all done through terminal commands, delivered in 20 minutes without opening Xcode.

In terms of permission control, each terminal command requires explicit user authorization before execution, providing two options: "Always Allow" and "Allow Once." My Computer also supports utilizing the local GPU for machine learning training or large model inference. Users can remotely assign tasks from devices such as phones to their home computers, as long as the computer is on and Manus Desktop is running. This feature integrates Manus's existing scheduled tasks and custom Agents, allowing users to automate local routine tasks, such as cleaning the downloads folder every morning or generating a local data summary report every week.

举报 Correction/Report
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish