BuyWhere

Posted on May 4

Build a Cross-Border Price Comparison Agent in 10 Minutes with BuyWhere MCP

#mcp #ai #tutorial #shopping

Ever tried to build an AI agent that searches real products across countries? Most agents hallucinate prices or hit rate limits. Here's how to get real-time product data from Singapore, SEA, and US markets in 3 lines of config.

What You'll Build

A Claude Desktop agent that can:

Search 50M+ real products across SEA and US markets
Compare prices across countries in real time
Never hallucinate product data again

What is BuyWhere MCP?

BuyWhere is a Model Context Protocol (MCP) server that gives AI agents access to a live product catalog spanning Lazada, Shopee, Amazon SG, and US merchants. It exposes 5 tools that any MCP-compatible client (Claude Desktop, Cursor, VS Code Copilot, Cline, OpenCode) can use.

The tools:

search_products — Natural language product search across markets
get_product — Full product details by ID
compare_products — Side-by-side comparison of 2-5 products
get_deals — Current price drops and promotions
list_categories — Category taxonomy

Step 1: Get Your Free API Key

Head to buywhere.ai/api-keys and sign up. The free tier includes 1,000 queries per month — no credit card required.

You'll get a key like bw_live_xxxx.

Step 2: Configure Your MCP Client

Claude Desktop: Add to claude_desktop_config.json:

{
  "mcpServers": {
    "buywhere": {
      "command": "npx",
      "args": ["-y", "@buywhere/mcp-server"],
      "env": {
        "BUYWHERE_API_KEY": "bw_live_xxxx"
      }
    }
  }
}

Cursor / VS Code / Cline: Same config, just put it in your MCP settings file (~/.cursor/mcp.json or VS Code's mcp.json).

OpenCode: Add to your opencode.json:

{
  "mcpServers": {
    "buywhere": {
      "command": "npx",
      "args": ["-y", "@buywhere/mcp-server"],
      "env": {
        "BUYWHERE_API_KEY": "bw_live_xxxx"
      }
    }
  }
}

Step 3: Start Searching

Restart your client. Now you can ask:

"Search for iPhone 16 Pro prices on Singapore vs US markets"

Your agent will call search_products and return real prices from live merchants:

Sony WH-1000XM5 (Singapore) — $398.00 @ Lazada SG
Sony WH-1000XM5 (Malaysia)  — $372.00 @ Shopee MY  
Sony WH-1000XM5 (US)        — $329.99 @ Amazon US

One query. Cross-border prices. No hallucination.

More Examples

Deal hunting:

"Find electronics with at least 30% off"

The agent calls get_deals(category="electronics", min_discount=30) and returns active promotions.

Gift recommendations:

"Find a birthday gift for a coffee-loving developer, budget $50"

The agent searches the catalog, cross-references deals, and returns curated picks with reasoning.

Inventory research:

"Find all gaming monitors under $400 from LG, Samsung, Dell"

Returns structured data for reseller analysis.

Why MCP Matters for Shopping

The Model Context Protocol is changing how AI agents access real-world data. Instead of LLMs guessing prices or scraping websites, agents use standardized MCP tools to fetch live, structured data.

BuyWhere fills a specific gap: cross-border product data for AI agents. No other MCP server gives agents access to Southeast Asian markets alongside US pricing in a single tool call.

What Makes This Different

Real data, not hallucination — every result comes from a live product catalog
Cross-border by default — search Singapore, Malaysia, and US in one query
MCP-native — works with any MCP client out of the box
Zero infrastructure — just npx and an API key

DEV Community