MCP OpenVision
by Nazruden
MCP OpenVision is a Model Context Protocol (MCP) server that provides image analysis capabilities powered by OpenRouter vision models. It enables AI assistants to analyze images via a simple interface within the MCP ecosystem.
Last updated: N/A
What is MCP OpenVision?
MCP OpenVision is an MCP server that allows AI assistants to analyze images using OpenRouter's vision models. It provides a simple interface for image analysis within the MCP ecosystem.
How to use MCP OpenVision?
MCP OpenVision can be installed via Smithery, pip, or UV. It requires an OpenRouter API key and can be configured through environment variables. You can test it with MCP Inspector or integrate it with Claude Desktop or Cursor by editing the MCP configuration file. The image_analysis tool is the core function, accepting image URLs, local file paths, or base64-encoded strings as input.
Key features of MCP OpenVision
image_analysis tool
Supports Base64, URLs, and file paths for image input
Configurable with OpenRouter API key
Integration with MCP ecosystem
Supports various vision models through OpenRouter
Use cases of MCP OpenVision
Analyzing images for retail product identification and price estimation
Analyzing medical scans for abnormalities
Extracting data from charts and identifying trends
Transcribing text from images like restaurant menus
FAQ from MCP OpenVision
What is the default vision model?
What is the default vision model?
The default model is qwen/qwen2.5-vl-32b-instruct:free, but you can specify any other compatible model.
How do I provide my OpenRouter API key?
How do I provide my OpenRouter API key?
You can set the OPENROUTER_API_KEY environment variable.
What image input types are supported?
What image input types are supported?
MCP OpenVision supports Base64-encoded strings, Image URLs (http/https), and File paths.
How do I use relative file paths?
How do I use relative file paths?
You can either ensure the path is relative to the current working directory or specify a project_root parameter.
How do I customize the image analysis?
How do I customize the image analysis?
You can use the query parameter to provide context and instructions, and the system_prompt to define the model's role.