What is the OCR MCP Server?

The OCR MCP Server is a MetaCall Protocol (MCP) server that utilizes macOS’s built-in Vision framework to perform Optical Character Recognition (OCR) on images.

What is MetaCall Protocol (MCP)?

MCP is an open protocol that standardizes how applications provide context to LLMs. MCP server acts as a bridge, allowing AI models to access and interact with external data sources and tools.

What are the key dependencies for the OCR MCP Server?

The key dependencies are Python 3.13+, ocrmac, Pillow, and mcp[cli]>=1.7.1.

How do I install the dependencies?

It is recommended to use a virtual environment. Create one with python -m venv .venv, activate it with source .venv/bin/activate, and then install the dependencies using uv sync.

How do I run the OCR MCP Server?

After installing the dependencies, run uv run main.py in your terminal.

What is the ocr_image tool?

The ocr_image tool is the core functionality of the OCR MCP Server. It takes an image file path as input and returns the recognized text segments, their confidence scores, and bounding box coordinates.

What input does the ocr_image tool require?

The ocr_image tool requires the file_path: str parameter, which specifies the absolute or relative path to the image file.

What output does the ocr_image tool provide on success?

On success, the ocr_image tool returns a JSON object containing the filename and an array of annotations. Each annotation includes the recognized text, confidence score, and bounding box coordinates.

What output does the ocr_image tool provide on error?

On error, the ocr_image tool returns a JSON object containing an error field with a descriptive error message.

Can I use the OCR MCP Server on non-macOS systems?

No, the OCR MCP Server relies on macOS’s Vision framework and will only function correctly on a macOS system.

How can I test the OCR MCP Server?

You can use the MCP Inspector to connect to the running MCP server and test the ocr_image tool.

How do I configure the OCR MCP Server in Cursor?

You can add a configuration block to your MCP JSON configuration file (~/.cursor/mcp.json or project-specific .cursor/mcp.json) that specifies the command and arguments to start the server.

Can I use the OCR MCP Server with the UBOS platform?

Yes, the OCR MCP Server is designed to integrate seamlessly with the UBOS platform for AI Agent orchestration and data connectivity. UBOS is a full-stack AI Agent development platform that helps you orchestrate AI Agents, connect them with your enterprise data, build custom AI Agents with your LLM model and Multi-Agent Systems.