Ripe MCP Overview
Give AI assistants, chatbots, and applications direct, structured access to real-time lodging, content, and availability, all while respecting your commission structure and goals.
Why We Built an MCP
Travelers increasingly start planning in a conversation, not a search bar. They ask an AI assistant where to stay, what's nearby, and whether it's actually available.
The Ripe MCP Server meets that shift head-on. It exposes Ripe's lodging search, property content, and live availability through the Model Context Protocol (MCP), an open standard that lets AI assistants and agent frameworks query external data sources in the same predictable way. Connect once, and any MCP-compatible assistant can ask Ripe real questions and get back live, structured answers.
New to MCP?Model Context Protocol is an open standard for connecting AI applications to external tools and data. If your assistant or agent framework already supports MCP, connecting to Ripe is a matter of pointing it at one URL. See Getting Started below.
Why Build on It
- Reach travelers where they already plan. Your lodging data becomes queryable by any AI assistant your travelers are already talking to.
- Less integration work. No custom API wrappers. Each tool describes its own inputs and outputs, so an assistant (or a developer) can figure out what's available on its own.
- LLM-ready responses. Clean, predictable JSON built for language models. No scraping, no fragile parsing.
- Live, accurate availability. Real pricing and bookability are confirmed at query time, so what an assistant tells a traveler matches what they see at checkout.
- Built to evolve. The API is versioned from the start, so new capability doesn't break existing integrations.
Who It's For
- Agencies and development teams building AI-powered travel experiences on behalf of Ripe clients
- Third-party AI travel tools that want real-time lodging inventory in their recommendation or discovery flows
- Ripe DMO clients, whose inventory can surface in AI-powered travel tools with no extra lift on their end
What You Can Query
The server exposes five read-only tools. An AI assistant calls whichever it needs, in whatever order the conversation calls for.
| Tool | What it returns |
|---|---|
list_destinations | The directory of active destinations available to search, with location metadata. Useful for narrowing a broad or vague request |
search_properties | Properties matching a place, date range, guest count, budget, amenities, rating, or free-text "vibe" (e.g., "cozy," "walkable," "ski-in/ski-out") |
get_property_content | Descriptions, photos, amenities, address, and review summaries for specific properties |
get_property_availability | Live, confirmed pricing and bookability for one property over a specific stay window |
list_search_vocabulary | The exact amenity and property-type vocabulary the search recognizes, for precise filtering |
A note on accuracy
search_propertiesis built for fast comparison across many options and may return an indicative rate. Before presenting a price as bookable, confirm it withget_property_availability. This is true whether you're a person or an AI.
REST API or MCP?
Both run on the same underlying lodging data. The right choice depends on what you're building.
| Ripe REST API | Ripe MCP Server | |
|---|---|---|
| Best for | Custom applications: booking engines, dashboards, internal tools | AI assistants, chatbots, and autonomous agents |
| Consumer | Your own code, calling endpoints you choose | An LLM, discovering and calling tools on its own |
| Integration effort | You design request/response handling per use case | Self-describing. The assistant reads tool definitions and adapts |
| Scope today | Full read access, with transactional capability | Read-only: search, content, and availability |
| Typical use case | "Build our booking flow" or "power our internal reporting" | "Let an AI assistant answer trip-planning questions with live data" |
In short: if you're building something a person clicks through, use the REST API. If you're building something an AI assistant talks to on a traveler's behalf, use MCP. Many integrations use both, side by side.
Getting Started
- Get access. See Authentication for how to obtain and use your credentials.
- Connect. Point your MCP-compatible client or agent framework at the Ripe MCP Server endpoint.
- Start asking. Your assistant will discover the available tools and begin querying.
Need more control? The Ripe REST API is available for custom builds and transactional use cases. See the API Reference for endpoints, parameters, and example responses.
What's Coming
We're expanding the toolset over time, including deeper content coverage and, longer term, moving beyond search and information into full booking. If you're building today, design for read-only, informational use. Anything transactional should go through the REST API in the meantime.
FAQ
Is there a cost to use this?
No. The MCP server is free to connect to today. In the future, a charge may be introduced for higher limits or added functionality.
Which AI assistants or frameworks does it work with?
Any client or framework that supports the Model Context Protocol, including Claude, AI chatbots, and most agent frameworks that support remote MCP servers.
Can I book a reservation directly through the MCP server?
Not yet. It's read-only today: search, content, and availability. Every response includes a link to complete a booking on the live site.
How fresh is the data?
Search and content results are served efficiently, but pricing and availability are always confirmed live at query time, never from a cache.
Where do I ask questions or report an issue?
Reach out to your Ripe account team, or use the developer support channel linked at the top of this hub.
This page describes the current, publicly accessible Ripe MCP Server. Capabilities will expand over time.
Updated 12 days ago