On May 20, Tencent announced the official launch of its operating system-level AI assistant Marvis (Mavis), available for download on the official website (Marvis.qq.com) without requiring an invitation code. In addition to conventional AI Agent capabilities, Marvis provides more intelligent system and file management, safer end-cloud computing switching, lower barrier for Agent services, and more convenient cross-device collaboration, based on a deep understanding of computer systems, computing power, files, applications, and other information.

image.png

Differing from traditional AI conversation tools or single-application Agents, Marvis is centered around the "operating system level." It integrates the terminal system, files, applications, computing power, and cross-device connections into a single AI intermediate layer: users don't need to remember setup entry points, file paths, or software operation steps; they just need to state their goal in one sentence, and Marvis can understand the task, break it down into steps, call the corresponding Agent to execute, and hand over to the user for confirmation at key stages such as privacy, security, and payments.

Operating System Level Understanding: Making Computers Smarter

Marvis turns the entire computer into a "smart object" that can directly access the system level, achieving true closed-loop control from system settings to data processing. A single sentence can call, find, and adjust Windows system settings: users can directly ask about computer configuration, battery health, network status, or let it check if hardware can run a game smoothly. They can also adjust system settings, optimize startup items, and clean up redundant files through chat.

image.png

For novice users and elderly users, it lowers the threshold for system settings and troubleshooting; for office users, it transforms scattered configuration, retrieval, and processing tasks into a more convenient entry point.

On this basis, Marvis combines multiple local models to understand files, allowing users to search for file/image content, recognize text within images, and provide features like AI photo albums and AI document libraries based on dimensions such as people, content themes, holidays, and locations, helping users build a personalized knowledge base and solving the problem of not finding files when you don't remember the filename.

In terms of security boundaries, Marvis establishes an L2-level safety mechanism, enforcing a "hard inquiry": sensitive operations such as deleting or modifying files and adjusting core system configurations will first present an execution plan and request user confirmation; higher sensitivity actions such as payments must be completed by the user themselves.

Optional Side-Device Models: Data Stays on Your Computer, No Cloud Upload

Marvis offers two working modes: "Efficiency Mode" and "Privacy Mode," allowing different users to choose suitable options based on the nature of the task and hardware configuration. For daily tasks, Efficiency Mode uses a combination of device and cloud: complex problem understanding and planning can use cloud models, while file processing, execution actions, and local indexing are completed on the device side as much as possible to ensure faster response and more stable experience.

"Privacy Mode" uses only device-side models, ensuring all data parsing, image recognition, and conversations happen locally without uploading to the cloud, meeting scenarios with high requirements for data security, such as finance, legal, and human resources. Even without internet, it can still be used normally.

This device-cloud collaboration is not simply choosing between "local or cloud," but breaking tasks into different levels: tasks that can be completed locally are processed locally, and cloud capabilities are called only when necessary. Due to extensive preprocessing on the device side, Marvis significantly reduces cloud Token consumption in tasks like image search and file understanding, lowering usage costs and minimizing unnecessary data transfer.

image.png

Multiple Preloaded Agents: Get to Work Right Away

Most mainstream Agent products require users to set up workflows or configure skills manually before use. Marvis pre-installs a multi-Agent collaboration system at the factory.

An "AI team" of six Agents is available 24/7: the main Agent handles overall coordination, understands needs, breaks down tasks, and schedules other Agents to complete them; the File Agent acts as a digital asset manager, handling file searches, reading, editing, generation, and format conversion; the Computer Agent functions as a Windows system maintenance specialist, managing low-level configurations and automation; the App Agent is responsible for application operations, calling apps and Exe applications on the computer; the Browser Agent serves as a web interaction specialist, taking over web pages for interaction and data scraping; the Search Agent is an expert in web searching and information aggregation, quickly locating answers from public resources and providing key citations. Users can issue a single instruction, and multiple Agents can process in parallel, displaying progress and results through execution logs and output areas.

This "preloaded multi-Agent mode" lowers the user's learning curve. For example, users can ask Marvis to find the latest quarter's invoices and organize them into a table in chronological order; they can also let it monitor flight prices regularly, track updates on a specific website, or organize materials for multi-platform releases.

image.png

Visual Cross-Device Control: More Convenient Collaboration

Marvis has powerful cross-device capabilities, supporting multiple platforms including Windows/Mac/iOS/Android, with account synchronization across devices: users can directly view and take control of their computer screen on their phone, and even unlock and control the PC remotely while it is locked; tasks that can be done on the computer can also be done remotely via the phone, creating a personal computer that you can carry with you. Unlike most products that are limited to "remote commands," Marvis provides real-time visual control at the desktop level.