Vision Product Analyzer MCP

Real MCP server for Claude Desktop. Classify product images, inspect quality for defects, and analyze products comprehensively.

• Model: Qwen2.5-VL (Apache 2.0 License)
• Pricing: Pay-Per-Event ($0.05, $0.10, $0.10)
• Connect: Standby URL in Claude Desktop

Features

1. Classify Product

Classify product images and extract key information.

Tool: classify_product
Input: imageUrl (URL of product image)
Output:
  - category (main category)
  - subcategory
  - tags (array of up to 5 tags)
  - condition (new | used | refurbished | damaged)
  - brand_visible (boolean)
  - brand_name (detected brand or null)
Price: $0.05

Example:

{
  "category": "Electronics",
  "subcategory": "Smartphones",
  "tags": ["Apple", "iPhone", "Used", "Screen Protected"],
  "condition": "used",
  "brand_visible": true,
  "brand_name": "Apple"
}

2. Inspect Quality

Inspect product images for quality issues and defects.

Tool: inspect_quality
Input: imageUrl (URL of product image)
Output:
  - overall_quality (excellent | good | acceptable | poor | defective)
  - defects_found (array of defect objects)
    - type (scratch, dent, crack, etc.)
    - severity (minor | moderate | severe)
    - location (description of location)
  - quality_score (1-10)
  - recommendation (pass | review | reject)
Price: $0.10

Example:

{
  "overall_quality": "good",
  "defects_found": [
    {
      "type": "scratch",
      "severity": "minor",
      "location": "bottom left corner"
    }
  ],
  "quality_score": 8,
  "recommendation": "pass"
}

3. Analyze Product

Comprehensive product analysis covering description, features, market positioning, and pricing.

Tool: analyze_product
Input: imageUrl (URL of product image)
Output:
  - analysis (structured text analysis including)
    - Product Description
    - Key Features
    - Target Market
    - Estimated Price Range
    - Condition Assessment
    - Competitive Positioning
Price: $0.10

Example Output:

Product Description: Premium smartphone in excellent condition
Key Features: High-end processor, OLED display, 5G capable, 256GB storage
Target Market: Tech enthusiasts, professionals seeking premium devices
Estimated Price Range: $700-$900
Condition Assessment: Used but well-maintained
Competitive Positioning: Mid-range premium segment, strong competition from similar devices

Connection (Claude Desktop)

1. In Claude Desktop → Settings → Developer
2. Add this server:

   {
     "vision-product-analyzer-mcp": {
       "url": "https://ntriqpro--vision-product-analyzer-mcp.apify.actor/mcp",
       "type": "stdio"
     }
   }
   

3. Or use the Standby URL directly in MCP configuration

Usage in Claude

Classify this product image: https://example.com/product.jpg
Inspect this product for defects: https://example.com/item.jpg
Analyze this product comprehensively: https://example.com/good.jpg

Local Development

npm install
npm start
# Server runs on http://localhost:3000/mcp

Testing

# Local
curl -X POST http://localhost:3000/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/list"
  }'

# Cloud (Standby URL)
curl -X POST "https://ntriqpro--vision-product-analyzer-mcp.apify.actor/mcp?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/list"
  }'

Pricing

Billed per successful call. Errors do not incur charges.

Powered By

• Qwen2.5-VL - Apache 2.0 License - Multi-modal vision language model for product analysis
• MCP SDK - Model Context Protocol for Claude Desktop integration
• Apify - Serverless actor runtime with MCP support

License

Apache 2.0 (Qwen2.5-VL attribution)

Support

For issues or feedback, contact the development team.