Manus

Manus is a hosted general-purpose AI agent designed to complete multi-step tasks and return artifacts rather than only answer a prompt. Official documentation describes a cloud sandbox with internet access, a persistent file system, and the ability to install software or create task-specific tools. Its product surfaces cover research, data analysis, document and presentation creation, website building, and browser-based work. Manus can be used from its web experience and mobile app, while the desktop app adds a My Computer capability for approved interaction with local files, applications, and development tools.

The service uses a credit-based plan structure, including a Free plan and paid options documented on the official plans page. Credit consumption varies with task complexity, and website hosting or user-facing website features can have separate usage-based charges. The REST API provides programmatic task creation, file management, and result retrieval. Plan prices and included credits can vary by account, billing period, and product surface; the live Manus plans page is the authority. Manus is proprietary; no canonical public source repository or public software license was identified for the hosted product.

Top features

• Cloud sandbox: tasks run in a virtual computer with network access and persistent task files, reducing the need to supervise every individual step.
• Research and analysis: gather sources, compare information, process files or datasets, and package the findings into a structured result.
• Artifact creation: produce reports, presentations, spreadsheets, web pages, and other downloadable or shareable outputs.
• Website builder: create and edit a site with credit-based development usage and separately metered hosting-related services where applicable.
• Desktop access: the desktop application can bridge Manus to approved local files and applications through the documented My Computer feature.
• Team and API workflows: shared workspaces support collaborative use, while the REST API can trigger tasks and collect outputs from another system.

Use cases

• Research competitors or a market and produce a cited decision brief.
• Analyze uploaded tables or documents and turn the findings into a report or presentation.
• Draft and iterate on a small website, then evaluate the separate hosting and runtime charges before publication.
• Use the desktop app for a task that needs explicitly granted access to local materials.
• Integrate repeatable research or content tasks into an internal workflow through the API.