home / mcp / rapidocr mcp server
Provides an MCP server exposing OCR endpoints to process images from content or file paths for automated text extraction.
Configuration
View docs{
"mcpServers": {
"z4none-rapidocr-mcp": {
"command": "uvx",
"args": [
"rapidocr-mcp"
]
}
}
}The RapidOCR MCP Server exposes OCR capabilities as an MCP service, enabling clients to submit images for text extraction through simple endpoints that accept image content or file paths.
You interact with the RapidOCR MCP Server through your MCP client by calling two endpoints. Use ocr_by_content to send raw image data as base64 and receive a list of detected text pieces. Use ocr_by_path to point to an image file on the server or accessible filesystem and obtain the same text results.
Prerequisites: ensure you have the required MCP runtime installed on your system. You will run the server using the runtime command shown in the usage example.
1) Start the RapidOCR MCP Server by running the following command in your terminal.
uvx run rapidocr-mcpThe MCP configuration for this server specifies the runtime command and arguments you use to start the server. The configuration below is what you place in your MCP config file to run the server.
"rapidocr-mcp": {
"command": "uvx",
"args": [
"rapidocr-mcp"
],
"env": {}
}The server exposes two OCR methods: ocr_by_content and ocr_by_path. ocr_by_content accepts base64-encoded image data and returns a list of detected text. ocr_by_path accepts a filesystem path to an image and returns the same type of results.
Be aware that large language models are increasingly capable of recognizing text in images, which may influence how you design your OCR workflows.
Performs OCR on an image content provided as base64 data and returns a list of detected text content.
Performs OCR on an image file specified by path and returns a list of detected text content.