mcp-openvision
8
MCP OpenVision is a Model Context Protocol server that enables image analysis using OpenRouter vision models. It simplifies image evaluations for AI assistants and supports multiple modes of image input, allowing integrations with platforms like Claude Desktop.
Overview
MCP OpenVision is a Model Context Protocol (MCP) server that provides image analysis capabilities powered by OpenRouter vision models. It allows AI assistants to analyze images via a simple interface within the MCP ecosystem.
Features
- image_analysis: Capable of analyzing images with vision models. Parameters include: image input, user query, and model selection. Supports multiple image input types such as base64 strings, URLs, and file paths.
Installation
- Install via Smithery:
npx -y @smithery/cli install @Nazruden/mcp-openvision --client claude - Using pip:
pip install mcp-openvision - Using UV:
uv pip install mcp-openvision
Usage
- Test with MCP Inspector:
npx @modelcontextprotocol/inspector uvx mcp-openvision - Integrate with Claude Desktop or Cursor with a configuration file.
Development
- Clone repository:
git clone https://github.com/modelcontextprotocol/mcp-openvision.git - Run tests using
pytest - Code formatting with Black.