Swiggy lets users order food via ChatGPT, Gemini, and Claude using AI integration

Swiggy Introduces AI Integration Across Multiple Services

Swiggy has launched Model Context Protocol (MCP) integration across its food delivery, quick-commerce, and dining platforms, marking a shift toward conversational commerce in India's delivery ecosystem 1

1

. The integration allows users to order food through AI assistants including OpenAI's ChatGPT, Google Gemini, and Anthropic's Claude by issuing natural language commands rather than navigating through traditional app interfaces 2

2

.

Developed by Anthropic, the Model Context Protocol (MCP) serves as an open-source framework that enables AI chatbots to connect with third-party data hubs and perform actions on behalf of users 1

1

. According to Swiggy CTO Madhusudhan Rao, this approach reflects how users now make decisions: "India's convenience needs are deeply contextual... conversational commerce allows users to simply express what they want, when they want it" 2

2

.

Swiggy Instamart Becomes First Quick-Commerce Platform With MCP

Swiggy Instamart has become the first quick-commerce platform globally to adopt MCP, offering access to over 40,000 products through AI agents 2

2

5

5

. Users can browse and purchase groceries, daily essentials, and ingredients using simple prompts like "Order ingredients for Thai green curry" or "Get me ingredients for Thai green curry" 2

2

. The AI agent handles the entire workflow—from searching and comparing options to applying offers, placing orders, and enabling users to track deliveries 2

2

.

For Swiggy Dineout, the integration extends to dining reservations, where AI agents can fetch available time slots, apply offers, and book restaurant tables through single prompts 2

2

.

Setting Up the MCP Connection Requires Manual Configuration

Users must manually configure the AI Integration through a multi-step process that involves navigating to Settings, selecting Connectors, and adding custom connector URLs for each service 3

3

. Swiggy provides three separate URLs: https://mcp.swiggy.com/food for food delivery, https://mcp.swiggy.com/instamart for grocery delivery, and https://mcp.swiggy.com/dineout for dining reservations 3

3

. Once connected, users can issue commands like "order a biryani I would love to eat" or "find the best protein snack for you from Instamart which is low in calories" 3

3

.

This approach mirrors similar implementations by other Indian companies, including Zerodha, which recently announced that users can connect their Kite user accounts with Claude to gain portfolio insights 3

3

. The trend suggests a broader movement toward API integration that exposes backend systems directly to AI agents rather than relying on traditional app interfaces.

Implementation Challenges Surface in Early Testing

MediaNama's hands-on testing revealed that the MCP integration functions as a gated developer feature rather than a consumer-ready product 4

4

. Access requires switching ChatGPT into developer mode, manually adding MCP server URLs, and authenticating through Swiggy's OTP-based login flow 4

4

. The integration does not appear by default during normal use of AI tools, making it inaccessible to users unfamiliar with technical configuration 4

4

.

Once authenticated, the AI gained access to saved delivery addresses, nearby restaurant listings, and the ability to initiate checkout flows 4

4

. However, menu access proved inconsistent. In multiple cases involving large chains such as Third Wave Coffee and Starbucks, the system could identify restaurants but failed to load menus, preventing items from being added to carts 4

4

. This incomplete exposure of menu APIs to the MCP layer means that restaurants without accessible menus effectively disappear from the AI-driven ordering experience, even though they remain visible in the standard Swiggy app 4

4

.

What This Means for Conversational Commerce in India

Swiggy's adoption of Model Context Protocol (MCP) signals an industry shift where AI agents could replace traditional app navigation for routine transactions. The ability to order food through AI assistants using natural language commands reduces cognitive load and streamlines decision-making for users juggling multiple daily tasks. For Google, OpenAI, and Anthropic, this integration validates their platforms as actionable commerce layers rather than just information retrieval tools.

Yet the current implementation raises questions about scalability and user adoption. The requirement for developer mode access and manual server configuration limits the feature to technically proficient users, while inconsistent menu access undermines reliability. If Swiggy intends to position conversational commerce as a mainstream interface, it will need to simplify onboarding and stabilize backend API integration to ensure consistent performance across all restaurants and product categories. Watch for updates on whether Swiggy transitions this from a gated developer feature to a one-click consumer experience, and whether competitors in India's crowded food delivery platform market follow suit with their own AI agent integrations.