What Is Model Context Protocol (MCP)?

Model Context Protocol (MCP) is an open-source standard for connecting AI applications to external systems. Using MCP, AI apps such as Claude or ChatGPT can connect to data sources such as databases, tools such as search engines, and workflows such as specialized prompts, which lets them access key information and perform tasks (What is the Model Context Protocol (MCP)?, Model Context Protocol).

Last updated: November 2026

How Model Context Protocol works

MCP uses a client-server design with three roles. The host is the AI application a person works in, and it opens one MCP client for each MCP server it connects to (Architecture overview, Model Context Protocol).

Before using a server, the client asks what it offers with a list request such as tools/list, then runs the tool it needs with tools/call. The client fetches that list at run time, so a server can change what it offers and connected applications see the new list the next time they ask (Architecture overview, Model Context Protocol).

For a hotel system, a tool might check availability or modify a booking, the two example tasks Apaleo's CTO gave when the company launched its MCP server (Apaleo launches MCP server, Apaleo).

Messages between client and server follow JSON-RPC 2.0. A remote server uses Streamable HTTP, where the client sends messages by HTTP POST and the server can stream replies with Server-Sent Events, and MCP recommends OAuth for obtaining authentication tokens (Architecture overview, Model Context Protocol).

The protocol is versioned by date. This page follows version 2026-07-28, the one in the current documentation, so some details differ from explainers written for earlier versions (Architecture overview, Model Context Protocol).

Why it matters for hotel distribution

Before MCP, connecting an AI agent to an external system took custom code for every pairing. With MCP, a system provider builds one server and an AI application builds one client, and any two that support the protocol can connect (The Protocol That's Reshaping How AI Talks to Travellers, OAG).

Apaleo launched an MCP server for its property management platform in September 2025 so AI agents can carry out cross-system hotel operations without weeks of custom coding, and it describes itself as the first property management platform in hospitality to do so (Apaleo launches MCP server, Apaleo). Use of Apaleo's server rose 70% between July and September 2026, by the company's figures, with hotels on its platform using agents for bulk reservation changes, forecasting and rate management (Hotels are putting AI agents to work in live operations, with Apaleo MCP Server use up 70%, Hospitality Net).

On the demand side, Expedia Group is preparing a B2B MCP server so partners' AI agents can connect directly to its travel inventory, and in May 2026 the company said it would go live in the coming months (Expedia to Launch Agentic AI Tools for B2B Partners, Skift). Earlier, in October 2025, Expedia, Booking.com and Tripadvisor launched AI integrations inside ChatGPT, used mostly for discovery and comparison, with bookings completed on their own sites (The Protocol That's Reshaping How AI Talks to Travellers, OAG).

Most hotel groups have not added this layer yet. Fifteen percent of hotel chains have a standardized AI data access layer such as MCP in place, and 52% have no plans to add one, in the h2c Hotel Chain Tech Report 2026 as cited by Hospitality Net (Hotels are putting AI agents to work in live operations, with Apaleo MCP Server use up 70%, Hospitality Net).

Model Context Protocol compared with a system's API

A hotel system's own API, such as a PMS, booking engine or channel connection, links other software to that one system. MCP links AI applications to external systems in general. An MCP server can sit on top of a system's API, and AI applications discover its tools at run time (Architecture overview, Model Context Protocol).

Frequently asked questions

What is Model Context Protocol in simple terms?
MCP works like a USB-C port for AI applications, the comparison its own documentation uses. USB-C gives electronic devices one standard way to connect, and MCP gives AI applications one standard way to connect to external systems. For a hotel, the property management platform is one such system once it offers an MCP server.
What is a Model Context Protocol server?
An MCP server is a program that provides context to MCP clients, and it can run on the same machine as the AI application or remotely. A server offers three kinds of building blocks: tools, which perform actions such as API calls and database queries; resources, which supply data; and prompts, which are reusable templates that help structure work with a language model.
Who created Model Context Protocol?
Anthropic introduced Model Context Protocol on November 25, 2024, as an open standard that lets developers build secure, two-way connections. It released the specification and SDKs at launch. On December 9, 2025, Anthropic donated MCP to the Agentic AI Foundation, a directed fund under the Linux Foundation that it co-founded with Block and OpenAI.
Who supports Model Context Protocol?
By December 2025 there were more than 10,000 active public MCP servers and more than 97 million monthly SDK downloads across Python and TypeScript, by the count of the company that created the protocol. ChatGPT, Cursor, Gemini, Microsoft Copilot and Visual Studio Code are among the AI products that had adopted it by then.
Is Model Context Protocol secure?
MCP cannot enforce security principles at the protocol level, so safety depends on how each application implements it. The specification treats tools as arbitrary code execution that must be handled with caution, and hosts must get the user's explicit consent before invoking any tool. For a property, the permissions each connected AI application holds are the thing to review before it can reach a system that stores bookings.

Where Jetstream fits

An AI agent connected through MCP works with the rooms, rates and availability that the connected systems hold, so those systems need to be right. Jetstream builds an AI channel manager that keeps a property's rooms, rates and availability accurate on every channel it sells through, managed from one place.