PaddleOCR Official MCP Server Documentation

PaddleOCR provides a lightweight MCP server, which can be used to quickly integrate PaddleOCR’s text recognition, document parsing, and other capabilities into various large model applications. The PaddleOCR MCP server currently supports the following tools:

• Currently Supported Tools

  • OCR: Perform text detection and recognition on images and PDF files.
  • PP-StructureV3: Identify and extract text blocks, titles, paragraphs, images, tables, and other layout elements from images or PDF files, converting the input into a Markdown document.

The PaddleOCR MCP server can be integrated into various applications to enable text recognition and document parsing capabilities. Here, we use Claude for Desktop as an example to demonstrate the steps for quickly connecting to the PaddleOCR MCP server:

1. Install uv.

2. Locate the Claude for Desktop configuration file in one of the following locations:

   • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%\Claude\claude_desktop_config.json
   • Linux: ~/.config/Claude/claude_desktop_config.json

   Open the claude_desktop_config.json file, refer to the following example to adjust the configuration, and paste it into your claude_desktop_config.json.

   For the OCR service:

   {
     "mcpServers": {
       "paddleocr-ocr": {
         "command": "uvx",
         "args": [
           "--from",
           "paddleocr-mcp@https://paddle-model-ecology.bj.bcebos.com/paddlex/PaddleX3.0/mcp/paddleocr_mcp/releases/v0.2.0/paddleocr_mcp-0.2.0-py3-none-any.whl",
           "paddleocr_mcp"
         ],
         "env": {
           "PADDLEOCR_MCP_PIPELINE": "OCR",
           "PADDLEOCR_MCP_PPOCR_SOURCE": "aistudio",
           "PADDLEOCR_MCP_SERVER_URL": "<your-server-url>",
           "PADDLEOCR_MCP_AISTUDIO_ACCESS_TOKEN": "<your-access-token>"
         }
       }
     }
   }

   For the PP-StructureV3 service:

   {
     "mcpServers": {
       "paddleocr-ocr": {
         "command": "uvx",
         "args": [
           "--from",
           "paddleocr-mcp@https://paddle-model-ecology.bj.bcebos.com/paddlex/PaddleX3.0/mcp/paddleocr_mcp/releases/v0.2.0/paddleocr_mcp-0.2.0-py3-none-any.whl",
           "paddleocr_mcp"
         ],
         "env": {
           "PADDLEOCR_MCP_PIPELINE": "PP-StructureV3",
           "PADDLEOCR_MCP_PPOCR_SOURCE": "aistudio",
           "PADDLEOCR_MCP_SERVER_URL": "<your-server-url>",
           "PADDLEOCR_MCP_AISTUDIO_ACCESS_TOKEN": "<your-access-token>"
         }
       }
     }
   }

   Obtain your access token from this page.

3. Restart Claude for Desktop. The new paddleocr-ocr tool should now be available in the application.

Besides the usage described above, the PaddleOCR MCP server also supports more features, such as integrating local Python libraries and using Streamable HTTP transmission. For more capabilities of the PaddleOCR MCP server, please refer to the official PaddleOCR documentation.