Desktop AI Client · Windows + Mac

Give Cherry Studio Desktop Access to Browser MCP Tools

Cherry Studio is a multi-model desktop AI client for Windows and macOS with native MCP support. Connect Agent MCP Studio's relay bridge to extend Cherry Studio with browser-built Python, SQL, and API tools — accessible from your desktop AI in minutes.

Share:

Prerequisites

Step-by-Step Setup

  1. Get your relay URL

    Open the studio → SettingsMCP Relay Bridge → copy URL → click Connect. Download bridge.js:

    curl -O https://agentmcp.studio/bridge.js && npm install ws
  2. Open Cherry Studio Settings → MCP

    In Cherry Studio: SettingsMCP ServerAdd MCP Server.

  3. Configure the MCP server

    {
      "name": "Agent MCP Studio",
      "type": "stdio",
      "command": "node",
      "args": [
        "C:\\path\\to\\bridge.js",
        "wss://agentmcp.studio/api/relay/YOUR-UUID"
      ]
    }

    On Windows, use double backslashes in paths or forward slashes.

  4. Enable tools in Cherry Studio chat

    Back in the chat view, toggle MCP Tools on from the toolbar. Your browser-built tools will appear. The model can call them automatically when relevant to your conversation.

Frequently Asked Questions

Yes — Cherry Studio has MCP support in its Settings → MCP Tools section. Add your bridge.js server configuration and Cherry Studio's AI chat gains access to all your Agent MCP Studio tools.

Cherry Studio is a feature-rich AI desktop client that supports multiple LLM providers (OpenAI, Anthropic, Google, local models). Popular in Asia, it offers a clean chat interface with knowledge bases, plugins, and MCP tool support.

Yes — Cherry Studio supports Alibaba Qwen, DeepSeek, Moonshot, Baidu ERNIE, and other Chinese LLMs alongside OpenAI and Anthropic. MCP tool calling works with any of these models that support function calling.

Cherry Studio is free and open-source (MIT license). Download it from GitHub. You supply your own API keys for cloud LLM providers, or configure local Ollama models at no cost.

Check: (1) MCP server is enabled in Cherry Studio settings, (2) the model you're using supports function calling, (3) relay UUID matches Agent MCP Studio Settings, (4) Node.js is installed, (5) the bridge.js absolute path is correct.

Related Integrations