Florence-2 MCP Server

Florence-2 MCP Server

An MCP server for processing images using Florence-2.

You can process images or PDF files stored on a local or web server to extract text using OCR (Optical Character Recognition) or generate descriptive captions summarizing the content of the images.

Installation

Claude

Download the latest MCP bundle mcp-florence2.mcpb from the Releases page, then open the downloaded .mcpb file or drag it into the Claude Desktop's Settings window.

{
  "mcpServers": {
    "florence-2": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/jkawamoto/mcp-florence2",
        "mcp-florence2"
      ]
    }
  }
}

For more information, see: Connect to local MCP servers - Model Context Protocol.

goose

Open this link

goose://extension?cmd=uvx&arg=--from&arg=git%2Bhttps%3A%2F%2Fgithub.com%2Fjkawamoto%2Fmcp-florence2&arg=mcp-florence2&id=florence2&name=Florence-2&description=An%20MCP%20server%20for%20processing%20images%20using%20Florence-2

to launch the installer, then click "Yes" to confirm the installation.

extensions:
  florence2:
    name: Florence-2
    cmd: uvx
    args: [ --from, git+https://github.com/jkawamoto/mcp-florence2, mcp-florence2 ]
    enabled: true
    type: stdio

For more details on configuring MCP servers in Goose, refer to the documentation: Using Extensions | goose.

LM Studio

To configure this server for LM Studio, click the button below.

Tools

ocr

Process an image file or URL using OCR to extract text.

Arguments:

• src: A file path or URL to the image file that needs to be processed.

caption

Processes an image file and generates captions for the image.

Arguments:

• src: A file path or URL to the image file that needs to be processed.

process

Processes an image file with a custom prompt using the Florence-2 model.

Arguments:

• src: A file path or URL to the image file that needs to be processed.
• prompt: A custom prompt for the Florence-2 model.

Security Considerations

When processing image files from URLs (src parameter starting with http:// or https://), outbound HTTP requests are initiated directly from the server process using Python's requests library. In environments where untrusted users or prompts may supply URLs, this could potentially expose the server to Server-Side Request Forgery (SSRF) targeting internal network services or cloud metadata endpoints.

To mitigate SSRF risks, it is strongly recommended to route all outgoing HTTP(S) traffic through an SSRF-safe forward proxy (such as Smokescreen or Squid configured to block internal IP ranges).

Since requests automatically respects standard environment variables, proxy settings can be configured without modifying the application code:

• HTTP_PROXY / http_proxy: Proxy URL for HTTP requests.
• HTTPS_PROXY / https_proxy: Proxy URL for HTTPS requests.
• NO_PROXY / no_proxy: Comma-separated list of hosts to bypass the proxy.

Configuration Examples

Claude Desktop

Add the env block to claude_desktop_config.json:

{
  "mcpServers": {
    "florence-2": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/jkawamoto/mcp-florence2",
        "mcp-florence2"
      ],
      "env": {
        "HTTP_PROXY": "http://proxy.example.com:8080",
        "HTTPS_PROXY": "http://proxy.example.com:8080"
      }
    }
  }
}

Goose

Add the env block to ~/.config/goose/config.yaml:

extensions:
  florence2:
    name: Florence-2
    cmd: uvx
    args: [ --from, git+https://github.com/jkawamoto/mcp-florence2, mcp-florence2 ]
    env:
      HTTP_PROXY: "http://proxy.example.com:8080"
      HTTPS_PROXY: "http://proxy.example.com:8080"
    enabled: true
    type: stdio

License

This application is licensed under the MIT License. See the LICENSE file for more details.