RapidOCR MCP Server

@z4none

Basic information

About RapidOCR MCP Server

A MCP server based on RapidOCR, providing an easy-to-use OCR interface.

RapidOCR MCP Server config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "rapidocr-mcp": {
      "command": "uvx",
      "args": [
        "run",
        "rapidocr-mcp"
      ]
    }
  }
}

RapidOCR MCP Server tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

Fetch the live tool list by running this server in a temporary sandbox using the button above.

Overview

What is RapidOCR MCP Server?

A MCP (Model Context Protocol) server based on RapidOCR that provides an easy-to-use OCR (Optical Character Recognition) interface. It allows large language models to extract text from images by exposing two OCR methods.

How to use RapidOCR MCP Server?

Run the server with uvx run rapidocr-mcp. Configure it in your MCP client using the provided JSON block with command: "uvx" and args: ["rapidocr-mcp"]. No additional environment variables are needed.

Key features of RapidOCR MCP Server

Exposes two OCR methods: ocr_by_content and ocr_by_path.
Accepts base64-encoded image data or a local file path.
Returns a list of TextContent objects with recognized text.
Easy setup via uvx with no extra configuration.
Designed to integrate seamlessly with MCP-compatible clients.

Use cases of RapidOCR MCP Server

Extract text from an image file stored on the local filesystem.
Extract text from in-memory image data (base64) passed by an LLM.
Enable LLMs to read text from scanned documents or screenshots without multimodal vision.

Frequently asked questions

What methods does RapidOCR MCP Server provide?

It provides `ocr_by_content` (for base64-encoded image data) and `ocr_by_path` (for an image file path). Both return a list of recognized text.

How do I install or run RapidOCR MCP Server?

Run `uvx run rapidocr-mcp` in your terminal. No separate installation step is required—uvx handles it.

Why use this server when modern LLMs already have multimodal capabilities?

The README notes that large language models are becoming increasingly powerful and generally possess multimodal abilities, but this server offers a dedicated OCR interface for cases where you prefer a separate, focused tool.

Comments

More Other MCP servers

Codelf

unbug

A search tool helps dev to solve the naming things problem.

14.2K

Glasswarp

See and control a real Windows PC you own — from any MCP client, locally or remotely. Observe (UIA + screenshots), click/type/drag/scroll, launch apps, owner Live View. BYOH: your machine, your key.