Wraps OpenAI's vision models in an MCP interface so you can analyze images by URL. You get one tool, describe-image, which takes an image URL and returns a text description of what's in it. Useful when you need to add image understanding to Claude or other MCP clients without building your own vision pipeline. Requires an OpenAI API key with credits. Built in TypeScript, runs via npx. The setup is straightforward: add it to your MCP config with your API key in the env block and you're ready to send image URLs for analysis.
claude mcp add --transport stdio mcp-s-ai-image-recognition-mcp uvx image-recognition-mcp