EasyOCR MCP tool
Overview
EasyOCR MCP Overview
EasyOCR MCP is a Model Context Protocol (MCP) server that exposes Optical Character Recognition (OCR) capabilities using the EasyOCR library. It allows you to extract text from images via three flexible tools: processing images from base64 strings, local files, or URLs.
Key Features
- Multiple Input Methods: OCR images from base64, file paths, or URLs.
- Multi-language Support: Configure supported languages via the
EASYOCR_LANGUAGESenvironment variable. - Flexible Output: Choose between text-only or detailed results (with bounding boxes and confidence scores).
- Performance Optimized: Caches OCR readers for faster repeated requests.
- Native EasyOCR Output: Returns results in EasyOCR’s original format.
Available Tools
- ocr_image_base64: OCR from a base64-encoded image string.
- ocr_image_file: OCR from a local image file.
- ocr_image_url: OCR from an image URL.
Each tool supports parameters for output detail, paragraph detection, and text merging thresholds.
Usage Example
# Example output with detail=1:
[
([[189, 75], [469, 75], [469, 165], [189, 165]], '愚园路', 0.37),
([[86, 80], [134, 80], [134, 128], [86, 128]], '西', 0.40)
]
# Example output with detail=0:
['愚园路', '西', '东', '315', '309', 'Yuyuan Rd.', 'W', 'E']
Configuration
- Languages: Set
EASYOCR_LANGUAGES(e.g.,en,ch_tra,ja) in your environment or MCP config. - Installation: Requires Python, EasyOCR, and PyTorch (with optional GPU support).
- Running: Start the server with
uv run easyocr-mcp.pyor via your MCP orchestrator.
Supported Languages
Supports 80+ languages, including English (en), Chinese (ch_sim, ch_tra), Japanese (ja), Korean (ko), French (fr), German (de), Spanish (es), and
Server Config
{
"mcpServers": {
"easyocr-mcp": {
"command": "uv",
"args": [
"--directory",
"/path/to/your/project/easyocr-mcp",
"run",
"easyocr-mcp.py"
],
"env": {
"EASYOCR_LANGUAGES": "en,ch_tra,ja"
}
}
}
}Project Info
Created At
a year agoUpdated At
a year agoAuthor Name
WindoCStar
-Language
-License
-Recommend Servers
View AllMcp Server Chatsum
@chatmcp
summarize chat message
typescript
a year ago