BuyWhere

Posted on May 3

Build an AI Price Comparison Agent in 10 Minutes with BuyWhere MCP

#mcp #ai #agents #tutorial

Want to build an AI agent that can search real products, compare prices across retailers, and find the best deal? Here is how to do it in 10 minutes with BuyWhere MCP.

What You Will Build

An AI shopping agent that:

Searches products across 180+ categories
Compares prices from multiple retailers
Finds the best deal for any product

Works with Claude Desktop, Cursor, Continue.dev, or any MCP-compatible client.

Step 1: Configure the MCP Server

Add this to your MCP client config:

{
  "mcpServers": {
    "buywhere": {
      "url": "https://api.buywhere.ai/mcp"
    }
  }
}

Get your free API key at buywhere.ai.

Step 2: Search Products

Ask your AI agent:

Search for wireless headphones under $200

The agent calls search_products and returns real products with prices, ratings, and availability.

Step 3: Compare Prices

Find the best price for Sony WH-1000XM5

The agent calls find_best_price and returns price comparisons across retailers.

Step 4: Get Product Details

Tell me about the Apple MacBook Pro M4

The agent calls get_product and returns full specs, images, and pricing.

What Makes This Different

Real data — not mock or synthetic. Millions of live products.
One endpoint — search, compare, and discover through a single MCP server.
Works anywhere — Claude, Cursor, VS Code, custom agents, any MCP client.

Challenge: Build Something Cool

We are running the Build With BuyWhere challenge (May 5-19). Build an agent using 2+ BuyWhere tools and win:

1st: Featured placement + 12 months free API tier ($1,188)
2nd: 12 months free API tier
3rd: 6 months free API tier

Full challenge details

Quick Start Template

For Claude Desktop users: just paste this into your config file and start asking about products.

For developers: here is a minimal Python agent example:

import json
import urllib.request

BUYWHERE_MCP = "https://api.buywhere.ai/mcp"
API_KEY = "your-key-here"

def search_products(query, max_price=None):
    payload = {
        "jsonrpc": "2.0",
        "method": "tools/call",
        "params": {
            "name": "search_products",
            "arguments": {
                "query": query,
                "limit": 5
            }
        },
        "id": 1
    }
    if max_price:
        payload["params"]["arguments"]["max_price"] = max_price

    req = urllib.request.Request(
        BUYWHERE_MCP,
        data=json.dumps(payload).encode(),
        headers={"Content-Type": "application/json", "X-API-Key": API_KEY}
    )
    return json.loads(urllib.request.urlopen(req).read())

result = search_products("wireless headphones", max_price=200)
print(json.dumps(result, indent=2))

Go Build

Get your API key at buywhere.ai, drop in the MCP config, and start building. The challenge runs May 5-19 — plenty of time to build something impressive.

Built with BuyWhere MCP — the product catalog API for AI agents.

DEV Community