Recently, the Windows version of Marvis was updated. The new client has lowered the minimum system requirements from 6 cores and 16GB to 4 cores and 8GB, allowing more low-spec PC devices to install and use Marvis' core AI capabilities. At the same time, the update adds a task notification feature, enabling users to promptly learn about the status and progress of their tasks.

Optimize Operation Mode, Lower the Usage Barrier

As a personal AI assistant at the operating system level, Marvis can search files and images on your computer, and also provide unique features such as AI images and AI document libraries based on dimensions like people, content themes, holidays, and locations. It solves the common problem of "not remembering the file name and thus not finding the file." Behind this are multiple local small models, which also result in basic requirements for processor and memory performance. In the early stages, the product set a threshold of 6 cores and 16GB primarily to address issues like resource competition and insufficient memory, ensuring stable operation of edge-side tasks.

In response to numerous user requests, Marvis has now lowered the usage barrier by using graded loading and low-resource adaptation to rebalance functionality and performance, supporting operation on 4-core 8GB devices. The specific approach includes: defaulting to close high-memory-consuming features, allowing manual activation or low-resource mode; reducing animations in modules like Office and work logs, and adjusting image loading according to device capabilities to reduce runtime pressure.

Optimize Core Functionality, Add Multiple Reminders to Ensure Stability

On low-spec devices, there are still three differences in functionality: the local knowledge base defaults to building a title index, while content indexing needs to be manually activated, and the system will prompt users that it may consume high memory before activation, defaulting to low-resource mode. Core functions in modules like Office and work logs are retained, but animation effects are reduced, with full animation experience better suited for devices with 16GB or more memory. High-computing power features like local mode maintain original requirements, and 4-core 8GB devices currently do not support them.

Complementing these changes, when device memory is insufficient, Marvis will guide users to free up resources before continuing to use it. In key areas such as settings, local knowledge base, search, and conversation, the product will actively explain which features are available, why they are limited, and how to handle it, without users needing to compare configuration tables themselves.

In recent years, Tencent has continuously promoted the expansion of AI capabilities to a broader user base, from open-sourcing large models and offering AI features for free to individual users, to lowering hardware and operational barriers for product usage. AI accessibility has always been a key direction for its product implementation. This latest reduction in device requirements for the Windows version of Marvis follows the same approach — making AI not only run on a few high-performance devices, but also enter more ordinary users' real computers and daily scenarios.