Cherry Studio's new version launches MCP function, unlocking a new AI interaction experience

The new version of Cherry Studio brings the MCP function, which takes the AI interaction experience to a higher level.
Core content:
1. Introduction to the MCP function and its application in Cherry Studio
2. Comparative analysis of the two MCP service types, STDIO and SSE
3. Two convenient ways to add an MCP server
Cherry Studio has added the highly anticipated MCP function in its latest version, bringing everyone a smarter and more efficient AI interaction experience.
1. First experience with MCP function
(Model Context Protocol), which allows large language models (LLMs) to obtain real-time information by calling external tools and services, greatly expanding their capabilities. In the new version of Cherry Studio, this function is perfectly integrated, allowing AI to no longer be limited to "knowledge memory" but to be able to "act in real time."
Open the new version of Cherry Studio and enter the settings interface, you will see the eye-catching "MCP Server" option. Here, you will start your journey of exploring the MCP function. Currently, Cherry Studio supports two types of MCP services: STDIO (standard input/output) and SSE (server-sent events).
STDIO Type
If you choose the STDIO type, it means you can run the MCP service locally. The advantage of this type is that it can directly access local files and applications, and achieve deep integration of AI and local resources. For example, you can let AI directly read local documents and tables for data analysis and content extraction; or call local drawing software to generate corresponding images according to your description. However, it should be noted that the use of the STDIO type requires the Python and NodeJS environments to be configured in advance to ensure the stable operation of the service.
SSE Type
The SSE type MCP service runs on a remote server. Its biggest feature is that it is easy to configure and can be quickly used without the need to build a complex local environment. However, since it runs remotely, it cannot directly access local resources like STDIO. But in terms of obtaining cloud data and calling online APIs, the SSE type performs well and can provide you with rich external information support.
2. How to add an MCP server
Method 1: Search and add
In the MCP server setting interface of Cherry Studio, click the "Search" button, and you will be pleasantly surprised to find that a large number of MCP services are already listed. These services cover a variety of fields, from web content retrieval, file format conversion, to data analysis tools in specific fields, etc. You only need to find the one that meets your needs according to the service description, click "Add Server", and then in the MCP server setting interface that pops up, just click OK at the bottom to easily complete the addition. The whole process is simple and fast, and even a technical novice can quickly get started.
Method 2: Edit JSON and add
If the MCP service you are looking for is not in the search list, don't worry, there is another way. You can go to the homepage of the target MCP service and get its Json configuration information. It should be noted here that you should choose the configuration that specifies the use of "uv" or "uvx" (corresponding to the previously installed dependencies). Copy the obtained code as a whole, return to the configuration page of Cherry Studio, click "Edit JSON", and then replace the original "mcpServers": {} part with the copied code as a whole (note that it must be replaced as a whole, and no brackets or punctuation can be omitted). After completing the replacement, click OK below the edit box, and the server will be successfully added. If you have added multiple MCP services in various ways, when you click "Edit JSON" again, you will find that the configuration information of all MCP services is recorded here. Therefore, when adding a new service through this method, be sure to carefully observe the information structure to avoid destroying existing content, otherwise the added service may become invalid.
3. Practical application scenarios of MCP function
Information retrieval and organization
In daily work and study, we often need to obtain useful information from a large number of web pages and documents. Now, with the MCP function of Cherry Studio, you only need to give instructions to AI, and it can quickly access web pages and read local documents by calling the corresponding MCP service, accurately extract the required information, and organize and summarize it according to your requirements. For example, when writing an academic paper, you can let AI help you search for relevant academic websites, download papers, and extract key arguments and research results, which greatly saves the time of consulting materials.
Smart Office Assistant
Cherry Studio's MCP function can also turn AI into your smart office assistant. It can be integrated with your office software to achieve functions such as automatic document summary generation, email classification and organization, and schedule optimization. For example, when you receive a large number of emails, AI can identify the email content through the MCP service, mark important emails, and classify them according to subject, sender, etc., making your mailbox management orderly.
IV. Precautions for using the MCP function
Model selection
When using the MCP function, not all models are supported. Therefore, before starting the MCP service, be sure to select a model that supports function calls. Generally speaking, in the model list of Cherry Studio, you can click the small gear behind the model to edit it, and select the function call option in More Settings to ensure that the model can work with the MCP function.
Service Enablement Management
After enabling the MCP service, each operation will interact with the large model, which will consume a certain amount of tokens. In order to avoid wasting resources, it is recommended to only enable the MCP services currently needed. In the chat window, you can flexibly select the MCP server to enable by clicking the MCP server button, so as to meet the needs while reasonably controlling the cost.
The new version of Cherry Studio's MCP function undoubtedly opens a door to more powerful AI applications. It makes AI more intelligent and practical, and can be better integrated into our work and life. Update to the latest version and experience the surprises brought by the MCP function yourself!