drio
Open app

Florence-2 MCP Server

Source

An MCP server for processing images using Florence-2

Catalog onlyCatalog onlySTDIO

Overview

Florence-2 MCP Server is a server designed for processing images using the Florence-2 model, which is capable of performing various image processing tasks.

To use the Florence-2 MCP Server, you need to configure it with either Claude Desktop or Goose CLI by editing the respective configuration files to include the server settings. After configuration, restart the application to start using the server.

  • Optical Character Recognition (OCR) capabilities for image files. - Ability to generate detailed captions for images. - Supports processing of both local image files and image URLs.
  1. Extracting text from scanned documents using OCR.
  2. Generating captions for images in applications like social media or content management systems.
  3. Automating image processing tasks in various workflows.

Add to your AI client

Use these steps to connect Florence-2 MCP Server in Cursor, Claude, VS Code, and other MCP-compatible apps. The same JSON appears in the Use with menu above for one-click copy.

Cursor

Add this to your .cursor/mcp.json file in your project root, then restart Cursor.

.cursor/mcp.json

{
  "mcpServers": {
    "mcp-florence2-jkawamoto": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-mcp-florence2-jkawamoto"
      ]
    }
  }
}

Claude Desktop

Add this server entry to the mcpServers object in your Claude Desktop config, then restart the app.

~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows)

{
  "mcpServers": {
    "mcp-florence2-jkawamoto": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-mcp-florence2-jkawamoto"
      ]
    }
  }
}

Claude Code

Add this to your project's .mcp.json file. Claude Code will detect it automatically.

.mcp.json (project root)

{
  "mcpServers": {
    "mcp-florence2-jkawamoto": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-mcp-florence2-jkawamoto"
      ]
    }
  }
}

VS Code (Copilot)

Add this to your .vscode/mcp.json file. Requires the GitHub Copilot extension with MCP support enabled.

.vscode/mcp.json

{
  "servers": {
    "mcp-florence2-jkawamoto": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-mcp-florence2-jkawamoto"
      ]
    }
  }
}

Windsurf

Add this to your Windsurf MCP config file, then restart Windsurf.

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "mcp-florence2-jkawamoto": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-mcp-florence2-jkawamoto"
      ]
    }
  }
}

Cline

Open Cline settings, navigate to MCP Servers, and add this server configuration.

Cline MCP Settings (via UI)

{
  "mcpServers": {
    "mcp-florence2-jkawamoto": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-mcp-florence2-jkawamoto"
      ]
    }
  }
}

FAQ

What types of images can be processed?

The server can process various image formats as long as they are accessible via file paths or URLs.

Is there a limit to the number of images processed at once?

The server can handle multiple images, but performance may vary based on the server's configuration and resources.

Is the Florence-2 MCP Server free to use?

Yes, the server is open-source and available for free under the MIT License.7:["$","div",null,{"className":"container mx-auto flex flex-col gap-4","children":["$L26","$L27",["$","$L28",null,{"currentProject":{"id":1261,"uuid":"24aec87c-3977-4d40-939d-197551af688b","name":"mcp-florence2","title":"Florence-2 MCP Server","description":"An MCP server for processing images using Florence-2","avatar_url":"https://avatars.githubusercontent.com/u/1793096?v=4","created_at":"2025-02-20T14:22:41.199Z","updated_at":"2025-02-23T07:20:28.001Z","status":"created","author_name":"jkawamoto","author_avatar_url":"https://avatars.githubusercontent.com/u/1793096?v=4","tags":"python,florence-2,mcp-server","category":"research-and-data","is_featured":false,"sort":1,"url":"https://github.com/jkawamoto/mcp-florence2","target":"_self","content":"$29","summary":"$2a","img_url":"https://github.com/jkawamoto/mcp-florence2/actions/workflows/python-app.yaml/badge.svg","type":null,"metadata":"{\"star\":\"1\",\"license\":\"MIT license\",\"language\":\"Python\",\"is_official\":false,\"latest_commit_time\":\"2025-05-04 02:46:16\"}","user_uuid":null,"tools":null,"sse_url":null,"sse_provider":null,"sse_params":null,"is_official":false,"server_command":null,"server_params":null,"server_config":null,"allow_call":false,"is_innovation":false,"is_dxt":false,"dxt_manifest":null,"dxt_file_url":null,"is_audit":false},"randomProjects":[],"currentServerKey":"$undefined"}]]}]

Research and data