Connect Any AI to Real-Time Grocery Data with QuickCommerce MCP Server

Imagine asking your AI assistant: **"What's the cheapest Amul Butter near me right now?"** — and getting back live prices from BlinkIt, Zepto, Swiggy Instamart, BigBasket, DMart Ready, JioMart, and Minutes within seconds. No code, no browser tabs, no spreadsheets.

That's exactly what the **QuickCommerce MCP Server** makes possible. MCP — the Model Context Protocol — is an open standard that lets AI tools call external APIs as natural extensions of their reasoning. Instead of copying data from websites into your chat, the AI fetches it directly.

QuickCommerce API is the first grocery data provider to offer a hosted MCP server. You don't install anything. You don't write any code. You just paste a URL into your AI tool's config and start asking questions about grocery prices in plain English.

BlinkItZeptoSwiggyBigBasketDMartJioMartMinutes

What Is MCP and Why Does It Matter?

The **Model Context Protocol (MCP)** is an open standard created by Anthropic that lets AI models call external tools during a conversation. Think of it as "plugins for AI" — but standardised, secure, and supported across multiple clients.

Before MCP, if you wanted Claude or ChatGPT to access live data, you had to build custom integrations, write wrapper scripts, or copy-paste data manually. MCP changes this by defining a universal protocol: the AI tool discovers available tools from an MCP server, understands their parameters, and calls them automatically when a user's query requires live data.

Here's what makes QuickCommerce's MCP server special: **it's hosted**. Most MCP servers require you to install Python packages, clone repos, and run local processes. Ours runs at `https://api.quickcommerceapi.com/mcp` — you just point your AI tool to the URL and you're done.

AI Tools That Work with QuickCommerce MCP

Any tool that supports the MCP Streamable HTTP transport can connect to QuickCommerce. Here are the most popular ones:

Setup: 60 Seconds to Live Grocery Data in Your AI

You need two things: your QuickCommerce API key (free 50 credits on signup at quickcommerceapi.com/auth/signup) and the MCP config below. That's it.

Claude Desktop / Claude Code

Open your MCP settings file (`~/.claude.json` for Claude Code, or Settings → Developer → MCP Servers in Claude Desktop) and add:

{
  "mcpServers": {
    "quickcommerce": {
      "url": "https://api.quickcommerceapi.com/mcp",
      "headers": {
        "X-API-Key": "your-api-key-here"
      }
    }
  }
}

Cursor / Windsurf

In Cursor, go to **Settings → MCP Servers → Add Server**. In Windsurf, open **Settings → MCP**. Use the same config:

{
  "name": "QuickCommerce",
  "url": "https://api.quickcommerceapi.com/mcp",
  "headers": {
    "X-API-Key": "your-api-key-here"
  }
}

Local Installation (Optional)

If you prefer running the MCP server locally (for airgapped environments or custom modifications), install the Python package:

# Install
pip install quickcommerce-mcp

# Run with your API key
QUICKCOMMERCE_API_KEY=your-key quickcommerce-mcp

{
  "mcpServers": {
    "quickcommerce": {
      "command": "quickcommerce-mcp",
      "env": {
        "QUICKCOMMERCE_API_KEY": "your-api-key-here"
      }
    }
  }
}

The 7 Tools Your AI Gets Access To

When you connect to the QuickCommerce MCP server, your AI tool automatically discovers all 7 tools. It knows each tool's parameters, descriptions, and when to use them. Here's the full toolkit:

Real Conversations: What You Can Ask

The magic of MCP is that you don't need to know API parameters or endpoints. You just ask naturally. Here are real conversations that work out of the box:

Example 1: "Find Me the Cheapest Atta"

You: Compare 5kg Atta prices across all platforms near
     Sarjapur Road, Bangalore (560103)

Claude: I'll search for Atta across all 7 platforms using group_search.

[Calls group_search with query="Atta 5kg",
 platforms="BlinkIt,Zepto,Swiggy,BigBasket,DMart,JioMart,Minutes",
 lat=12.9021, lon=77.6639, pincode="560103"]

Here's what I found:

| Platform   | Product                        | Price  |
|------------|--------------------------------|--------|
| DMart      | Samrat MP Chakki Wheat Atta    | ₹206   |
| BlinkIt    | Whole Farm Chakki Atta         | ₹207   |
| JioMart    | Good Life Chakki Atta          | ₹209   |
| Swiggy     | Fortune Chakki Fresh Atta      | ₹227   |
| BigBasket  | Aashirvaad Shudh Chakki Atta   | ₹245   |
| Zepto      | Aashirvaad Atta 5kg            | ₹249   |

DMart has the best price at ₹206. BlinkIt is a close second
at ₹207 with faster delivery (8 minutes vs DMart's 2PM-5PM slot).

Example 2: "Who Delivers Fastest?"

You: Which platform can deliver to Andheri West, Mumbai
     the fastest right now?

Claude: Let me check delivery ETAs for all platforms.

[Calls group_eta with platforms="BlinkIt,Zepto,Swiggy,BigBasket,DMart,JioMart,Minutes",
 lat=19.1364, lon=72.8296, pincode="400058"]

Here are the current delivery times:

| Platform   | ETA        | Status     |
|------------|------------|------------|
| Zepto      | 8 minutes  | ✅ Open    |
| BigBasket  | 10 minutes | ✅ Open    |
| BlinkIt    | 12 minutes | ✅ Open    |
| Swiggy     | 15 minutes | ✅ Open    |
| Minutes    | 22 minutes | ✅ Open    |
| JioMart    | N/A        | 🕐 Slot-based |
| DMart      | 2PM-5PM    | 🕐 Slot-based |

Zepto is fastest at 8 minutes. If you need it NOW, go with Zepto.
For bulk orders, DMart's 2PM-5PM slot has better prices.

Example 3: "Plan My Weekly Grocery List"

You: I need milk, bread, eggs, and onions near HSR Layout,
     Bangalore. Find me the cheapest option for each across
     all platforms.

Claude: I'll search for each item across all platforms and find
the best deal for each.

[Calls group_search 4 times — once per item]

Here's your optimized shopping list:

| Item       | Best Platform | Price  | Delivery |
|------------|--------------|--------|----------|
| Amul Milk  | DMart        | ₹27    | 2PM-5PM  |
| White Bread| BlinkIt      | ₹40    | 10 min   |
| Eggs (12)  | Zepto        | ₹72    | 8 min    |
| Onion 1kg  | BigBasket    | ₹35    | 12 min   |

Total: ₹174 if you split across platforms
vs ₹198 on BlinkIt alone (saving ₹24)

Quick tip: If convenience matters more, BlinkIt has everything
at ₹198 delivered in 10 minutes.

Example 4: "Competitive Price Audit" (Business)

You: I'm a brand manager for XYZ Ghee. Search for "ghee" on
     all platforms in Mumbai and show me how our product is
     priced compared to competitors like Amul and Gowardhan.

Claude: I'll do a competitive price audit across all platforms.

[Calls group_search with query="ghee", all 7 platforms, Mumbai coords]

Competitive Price Audit — Ghee (Mumbai 400001):

| Brand         | BlinkIt | Zepto  | Swiggy | BigBasket | DMart  |
|---------------|---------|--------|--------|-----------|--------|
| Amul Ghee 1L  | ₹620    | ₹615   | ₹625   | ₹610      | ₹599   |
| Gowardhan 1L  | ₹545    | ₹550   | ₹549   | ₹540      | ₹529   |
| XYZ Ghee 1L   | ₹580    | ₹575   | N/A    | ₹570      | N/A    |

Key findings:
- Your product is NOT listed on Swiggy and DMart — missing revenue
- You're priced 6% higher than Gowardhan on BlinkIt
- BigBasket offers your most competitive placement at ₹570

Under the Hood: How the MCP Server Works

The QuickCommerce MCP Server uses the **Streamable HTTP transport** — the newest and most efficient MCP transport. Unlike older SSE-based servers, Streamable HTTP uses standard HTTP POST requests with JSON-RPC, making it compatible with any HTTP client and firewall-friendly.

POST https://api.quickcommerceapi.com/mcp
Content-Type: application/json
Accept: application/json, text/event-stream
X-API-Key: your-api-key

{
  "jsonrpc": "2.0",
  "method": "initialize",
  "id": 1,
  "params": {
    "protocolVersion": "2025-03-26",
    "capabilities": {},
    "clientInfo": {
      "name": "claude-desktop",
      "version": "1.0"
    }
  }
}

{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "protocolVersion": "2025-03-26",
    "capabilities": {
      "tools": { "listChanged": false }
    },
    "serverInfo": {
      "name": "QuickCommerce API",
      "version": "1.26.0"
    }
  }
}

POST https://api.quickcommerceapi.com/mcp
Content-Type: application/json
X-API-Key: your-api-key

{
  "jsonrpc": "2.0",
  "method": "tools/call",
  "id": 3,
  "params": {
    "name": "group_search",
    "arguments": {
      "query": "Amul Butter 500g",
      "platforms": "BlinkIt,Zepto,Swiggy",
      "lat": 12.9021,
      "lon": 77.6639,
      "api_key": "your-api-key"
    }
  }
}

{
  "jsonrpc": "2.0",
  "id": 3,
  "result": {
    "content": [
      {
        "type": "text",
        "text": "{\n  \"BlinkIt\": {\n    \"products\": [\n      {\n        \"name\": \"Amul Pasteurised Butter\",\n        \"brand\": \"Amul\",\n        \"price\": 295,\n        \"mrp\": 310,\n        \"quantity\": \"500 g\"\n      }\n    ]\n  },\n  \"Zepto\": { ... },\n  \"Swiggy\": { ... }\n}"
      }
    ]
  }
}

Hosted vs Local: Which Should You Choose?

Credit Costs: What Does It Cost?

MCP calls use the same credit system as the REST API. The AI tool deducts credits for each tool call it makes. Here's the breakdown:

A typical question like "compare Atta on all 7 platforms" costs 7 credits (one per platform). You can ask "how many credits do I have?" anytime — it's free. Start with 50 free credits on signup, then top up from ₹99 for 500 credits.

Try It Without Installing Anything

Don't have Claude Desktop or Cursor? You can try the MCP-powered chat directly in your browser at quickcommerceapi.com/mcp/chat. It uses the same tools, shows the same rich product cards, and works on mobile too. Just log in and start asking.

Security & Privacy

Your API key is sent via the `X-API-Key` header over HTTPS — it never appears in URLs or logs. The MCP server is stateless: it doesn't store your queries, results, or conversation history. Each request is authenticated, executed, and forgotten.

The hosted MCP server runs on the same AWS App Runner infrastructure as the REST API, in `ap-south-1` (Mumbai). All traffic is encrypted with TLS 1.3. The server has no access to your AI tool's conversation history — it only sees the specific tool call parameters.

Build Your Own Agent with MCP

The MCP server isn't just for chat interfaces. You can build autonomous agents that use QuickCommerce tools to make purchasing decisions, monitor prices, or generate reports. Here's a minimal Python example using the MCP SDK:

import asyncio
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client

async def price_check():
    """Check Amul Butter prices across platforms."""
    headers = {"X-API-Key": "your-api-key"}

    async with streamablehttp_client(
        "https://api.quickcommerceapi.com/mcp",
        headers=headers,
    ) as (read, write, _):
        async with ClientSession(read, write) as session:
            await session.initialize()

            # List available tools
            tools = await session.list_tools()
            print(f"Available tools: {[t.name for t in tools.tools]}")

            # Search across platforms
            result = await session.call_tool(
                "group_search",
                arguments={
                    "query": "Amul Butter 500g",
                    "platforms": "BlinkIt,Zepto,Swiggy,BigBasket",
                    "lat": 12.9021,
                    "lon": 77.6639,
                    "api_key": "your-api-key",
                },
            )

            print(result.content[0].text)

asyncio.run(price_check())

import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { StreamableHTTPClientTransport }
  from "@modelcontextprotocol/sdk/client/streamableHttp.js";

const transport = new StreamableHTTPClientTransport(
  new URL("https://api.quickcommerceapi.com/mcp"),
  { requestInit: { headers: { "X-API-Key": "your-api-key" } } }
);

const client = new Client({ name: "my-agent", version: "1.0" });
await client.connect(transport);

// Call any tool
const result = await client.callTool({
  name: "group_search",
  arguments: {
    query: "Milk 1L",
    platforms: "BlinkIt,Zepto",
    lat: 19.076,
    lon: 72.8777,
    api_key: "your-api-key",
  },
});

console.log(result.content);

Frequently Asked Questions

Does the AI store my grocery searches?

The QuickCommerce MCP server is stateless — it does not store queries or results. Your AI tool (Claude, Cursor, etc.) may store conversation history according to its own privacy policy, but the MCP server forgets everything after each request.

Can I use MCP with ChatGPT or Gemini?

Not yet — MCP is currently supported by Anthropic's Claude and several code editors. However, you can use the REST API with any AI tool that supports function calling (OpenAI, Gemini, etc.). The browser chat also works for everyone.

What happens if I run out of credits mid-conversation?

The tool will return an error message saying "Insufficient credits" with a link to the pricing page. The AI will relay this to you. You can top up instantly and continue the conversation.

Is pincode required for all platforms?

Pincode is only required for **DMart, JioMart, and Minutes**. For BlinkIt, Zepto, Swiggy, and BigBasket, latitude and longitude are sufficient. If you ask the AI about DMart without a pincode, it will tell you one is needed.

Get Started Now

The QuickCommerce MCP Server is live and ready. Sign up, grab your API key, paste the config, and ask your AI about grocery prices. It takes less than 60 seconds.