AI Vision

v1.0.0โ€ขData Science & MLโ€ขstable

๐Ÿ“‡ ๐Ÿ  ๐ŸŽ ๐ŸชŸ ๐Ÿง - Multimodal AI vision MCP server for image, video, and object detection analysis. Enables UI/UX evaluation, visual regression testing, and interface understanding using Google Gemini and Vertex AI.

ai-visionmcpai-integration
Share:
50
Stars
0
Downloads
0
Weekly
0/5

What is AI Vision?

AI Vision is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to ๐Ÿ“‡ ๐Ÿ  ๐ŸŽ ๐ŸชŸ ๐Ÿง - multimodal ai vision mcp server for image, video, and object detection analysis. enables ui/ux evaluation, visual regression testing, and interface understanding using google gemini a...

๐Ÿ“‡ ๐Ÿ  ๐ŸŽ ๐ŸชŸ ๐Ÿง - Multimodal AI vision MCP server for image, video, and object detection analysis. Enables UI/UX evaluation, visual regression testing, and interface understanding using Google Gemini and Vertex AI.

This server falls under the Data Science & ML category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • ๐Ÿ“‡ ๐Ÿ  ๐ŸŽ ๐ŸชŸ ๐Ÿง - Multimodal AI vision MCP server for image, video

Use Cases

Analyze images, videos, and objects using multimodal AI. Evaluate UI/UX, test for visual regressions, and understand interfaces.
tan-yong-sheng

Maintainer

LicenseMIT
Languagetypescript
Versionv1.0.0
UpdatedMay 8, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx ai-vision

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use AI Vision

AI Vision MCP Server is a multimodal analysis server that connects AI assistants to Google Gemini or Vertex AI vision models, enabling them to analyze images, compare multiple images, detect and annotate objects with bounding boxes, audit UI/UX designs for accessibility and contrast compliance, and analyze video content including YouTube URLs. It exposes five purpose-built tools covering the most common visual AI tasks, with configurable temperature and model settings per task type. Teams doing visual regression testing, design reviews, or content moderation can invoke these capabilities through natural language without writing any vision API code.

Prerequisites

  • Node.js 18 or later installed
  • A Google AI Studio account with a Gemini API key (GEMINI_API_KEY), OR a Google Cloud project with Vertex AI enabled and a service account with the necessary permissions
  • For Vertex AI: a GCS bucket for video uploads (set via GCS_BUCKET_NAME)
  • An MCP-compatible client such as Claude Desktop, Cursor, or Claude Code
1

Obtain your API credentials

For Google AI Studio: visit https://aistudio.google.com/app/apikey to generate a Gemini API key. For Vertex AI: create a service account in your GCP project, grant it the 'Vertex AI User' role, and download the JSON key to extract the client_email and private_key.

2

Add the server via Claude Code CLI (Google AI Studio)

The simplest setup uses Claude Code's mcp add command to register the server with your Gemini API key.

claude mcp add ai-vision-mcp \
  -e IMAGE_PROVIDER=google \
  -e VIDEO_PROVIDER=google \
  -e GEMINI_API_KEY=your-gemini-api-key \
  -- npx ai-vision-mcp
3

Or configure manually in claude_desktop_config.json

For Claude Desktop or any MCP client that uses a JSON config file, add the server entry with the required environment variables.

{
  "mcpServers": {
    "ai-vision-mcp": {
      "command": "npx",
      "args": ["ai-vision-mcp"],
      "env": {
        "IMAGE_PROVIDER": "google",
        "VIDEO_PROVIDER": "google",
        "GEMINI_API_KEY": "your-gemini-api-key"
      }
    }
  }
}
4

Configure for Vertex AI (optional alternative)

If you prefer Vertex AI for enterprise billing and data residency, use the Vertex AI provider configuration instead.

{
  "mcpServers": {
    "ai-vision-mcp": {
      "command": "npx",
      "args": ["ai-vision-mcp"],
      "env": {
        "IMAGE_PROVIDER": "vertex_ai",
        "VIDEO_PROVIDER": "vertex_ai",
        "VERTEX_CLIENT_EMAIL": "[email protected]",
        "VERTEX_PRIVATE_KEY": "-----BEGIN PRIVATE KEY-----\n...\n-----END PRIVATE KEY-----\n",
        "VERTEX_PROJECT_ID": "your-gcp-project-id",
        "GCS_BUCKET_NAME": "ai-vision-mcp-your-project-id"
      }
    }
  }
}
5

Restart your MCP client and verify the tools load

Restart Claude Desktop or reload your client. The five vision tools (analyze_image, compare_images, detect_objects_in_image, audit_design, analyze_video) should appear as available tools.

AI Vision Examples

Client configuration

Claude Desktop configuration using Google AI Studio (Gemini API key) as the vision provider.

{
  "mcpServers": {
    "ai-vision-mcp": {
      "command": "npx",
      "args": ["ai-vision-mcp"],
      "env": {
        "IMAGE_PROVIDER": "google",
        "VIDEO_PROVIDER": "google",
        "GEMINI_API_KEY": "your-gemini-api-key"
      }
    }
  }
}

Prompts to try

Use natural language to invoke vision analysis. Provide image paths (local or URL), YouTube links, or paths to design screenshots.

- "Analyze the image at https://example.com/screenshot.png and describe what you see"
- "Compare these two screenshots and tell me what changed: /tmp/before.png and /tmp/after.png"
- "Detect all objects in /tmp/product_photo.jpg and save an annotated version to /tmp/annotated.png"
- "Audit the UI design at /tmp/dashboard.png for WCAG contrast compliance and usability issues"
- "Analyze this YouTube video and summarize what happens: https://www.youtube.com/watch?v=abc123"

Troubleshooting AI Vision

API key error: 'GEMINI_API_KEY is not set' or '400 API key not valid'

Ensure the GEMINI_API_KEY environment variable is set correctly in your MCP client config. Verify the key is active in Google AI Studio (https://aistudio.google.com/app/apikey) and that the Generative Language API is enabled for your project.

Video analysis times out for long videos

Set longer MCP timeouts in your Claude Code settings: add `MCP_TIMEOUT=60000` and `MCP_TOOL_TIMEOUT=300000` (5 minutes) to your ~/.claude/settings.json env section. Large videos uploaded to GCS may take additional time to process.

Vertex AI authentication fails with 'invalid_grant' or 'UNAUTHENTICATED'

Ensure VERTEX_PRIVATE_KEY contains the full PEM-formatted key including BEGIN/END headers and that newlines are encoded as \n in the JSON config. The service account must have the 'Vertex AI User' (roles/aiplatform.user) IAM role on the project.

Frequently Asked Questions about AI Vision

What is AI Vision?

AI Vision is a Model Context Protocol (MCP) server that ๐Ÿ“‡ ๐Ÿ  ๐ŸŽ ๐ŸชŸ ๐Ÿง - multimodal ai vision mcp server for image, video, and object detection analysis. enables ui/ux evaluation, visual regression testing, and interface understanding using google gemini and vertex ai. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install AI Vision?

Follow the installation instructions on the AI Vision GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with AI Vision?

AI Vision works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is AI Vision free to use?

Yes, AI Vision is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Data Science & ML MCP Servers

Explore all data science & ml servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "ai-vision": { "command": "npx", "args": ["-y", "ai-vision"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide โ†’

Ready to use AI Vision?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides