AirDNA Adapt -notice

Meet AirDNA Adapt. Pricing built on the data you already trust.

AirDNA Product Features

How to Automate Pricing Checks With Adapt MCP

Published: November 9, 2026

Last updated: November 13, 2026

Jessica Spinks
By

Jessica Spinks, CRME

Automate Pricing Checks with Adapt MCP

At a glance

  • AirDNA Adapt now has an MCP server that connects Claude and other AI assistants to your Adapt account
  • It reads your listings, calendars, reservations, pricing rules, comps and pacing, and can change rules, strategies, seasons and price tiers
  • Nothing gets pushed to your channels without a preview and your explicit yes
  • Setup takes about five minutes from the Claude connector directory
  • I walked through my own daily setup live on Thursday, November 12. Watch the recording here.

AirDNA Adapt now has an MCP server. That means you can connect Claude to your Adapt account and ask it, in plain English, how your portfolio is pacing, which units need attention this morning, and what to do about them. Then you can have it make the change.

After eight years in revenue management and nearly 15 in the short-term rental industry, I’ve learned that a pricing tool can show you what is happening. Understanding why, and acting before it becomes a problem, is where the revenue manager earns their keep. The hard part has never been the decision. It has been getting to the decision through the PMS, the pricing tool and a spreadsheet, one unit at a time.

That is what changed for me. Before every call with a property manager I used to sift. Now I open one dashboard that Claude built on top of the Adapt MCP, and it tells me which units picked up two bookings this week, which ones are pacing 20% behind last year, and which weekends are stuck behind a three-night minimum that nobody is going to book. The sifting is gone. What’s left is the strategy.

The payoff of getting pricing right is significant. Listings priced algorithmically earn about 19% more revenue per available night than comparable listings priced by hand. For a typical full-time home, that’s roughly $5,700 a year.¹ The MCP doesn’t replace that pricing engine. It gives you a way to watch it, question it, and steer it without clicking through 800 calendars.

This post covers what the Adapt MCP is, what it can read and change, how to connect it, what to ask it first, how I run my mornings with it, and how it keeps you in control.

AirDNA Adapt MCP

Try Adapt

Dynamic pricing that considers the entire competitive landscape before pricing a single night. Connect your listing to get started.

Try Adapt

What Is the AirDNA Adapt MCP?

MCP stands for Model Context Protocol. It is the standard way AI assistants like Claude plug into other software. You don’t need to know how it works under the hood. You connect it once, and from then on your assistant can see your Adapt data and use Adapt’s tools on your behalf.

AirDNA’s MCP server lives at mcp.airdna.co. It is in beta, it has public documentation, and it is listed in the Claude connector directory, so adding it is one click. It is built on the same data that powers everything else at AirDNA: 15M+ active short-term rental listings across 120K+ global markets, with data back to 2014. That matters because an assistant is only as useful as what it can see. Ask it about your comps and it is looking at real comps. Ask it how your listings sit against their market and it is looking at AirDNA’s market data, not guessing.

One thing it is not: a scraper. The “Airbnb MCP” projects you may have seen on GitHub are community tools for travelers that search public Airbnb pages. They don’t log into your host account and can’t touch your calendar. The Adapt MCP works on the listings you have connected to Adapt, with your login, and only yours.

Is there an official Airbnb MCP server?

No. Airbnb does not offer an MCP server or a public API for hosts, so there is no way to connect Claude to Airbnb directly. What exists are community search tools for guests, and MCP servers from the software you already use to run your listings. Adapt’s is the one that handles pricing.

What Can Claude Read and Change Through the Adapt MCP?

The short version: anything you can see in Adapt, your assistant can read, and most of what you can change in Adapt, it can change too, with guardrails. It works on the listings you have connected to Adapt; until a listing is connected, the assistant will tell you it can’t find any properties.

On the read side, the tools map to the questions you already ask every day. What does my calendar look like for the next 90 days, with price and minimum stay per night. What booked, when, through which channel, how far out, at what rate. How does my pacing compare with this day last year. What are my comps charging for the same dates, booked and still open.

AirDNA Adapt MCP Pricing Opportunities

On the write side, it can create or edit any pricing rule Adapt supports: day overrides, floors and ceilings, minimum stays, length-of-stay discounts, day-of-week pricing, gap-night rules, last-minute and far-out adjustments, seasons and events, and the price tier your listing is priced against. It can apply a strategy. And it can push previewed rates to your PMS, once you have approved the preview.

A note on privacy. The assistant only ever sees your own listings. It never returns guest names or contact details, it doesn’t return your listing’s street address, and comp properties are described by title and structure, not by ID.

There are around 30 tools live today, and the full list with every parameter is at mcp.airdna.co/docs.

How Do You Connect Claude to Adapt?

You need three things: a Claude account (web or desktop), an AirDNA login, and at least one listing connected to Adapt. If you haven’t connected a listing yet, do that first; the assistant will tell you it can’t find any properties until you do. Connecting works with Airbnb directly or through Guesty, Hostaway, Hospitable, OwnerRez and Uplisting.

Then:

  1. In Claude, go to Settings → Connectors → Add custom connector. Or skip the typing and use the one-click listing at claude.ai/directory/airdna.
  2. For the server URL, enter https://mcp.airdna.co
  3. Open Advanced settings and enter the OAuth Client ID: 478a2ed7-0d70-4feb-a340-d8c96dcd4c2b. This ID isn’t a secret; it’s the same for everyone and just tells Claude it’s talking to AirDNA.
  4. Save, then sign in with the same email and password you use for AirDNA.

To check it worked, start a new chat and ask: “What AirDNA tools do you have?” If nothing shows up, disconnect the AirDNA connector in Claude’s settings and connect again. That fixes it most of the time.

AirDNA Claude MCP server

Connect ChatGPT to Adapt

ChatGPT works too. The setup is a little more technical than Claude’s one-click listing, but it takes a few minutes and you only do it once:

  1. In ChatGPT, go to Plugins, then Add (+), then Add custom MCP server.
  2. Enter the connection details. Name: AirDNA. Description: Short-term rental market insights (or anything you like). Server URL: https://mcp.airdna.co. Authentication: OAuth.
  3. Open Advanced OAuth settings and select User-Defined OAuth Client. OAuth Client ID: 478a2ed7-0d70-4feb-a340-d8c96dcd4c2b. Leave the Client Secret blank and set the token endpoint auth method to none.
  4. Check “I understand and want to continue” and click Create as a plugin.
  5. Sign in with your AirDNA email and password, then start a chat and ask ChatGPT what AirDNA tools it has.

AirDNA Adapt ChatGPT MCP server

If you use Claude Code or Codex instead, the setup is one command each; the connection section of the docs has them. Support for Cursor, VS Code and other assistants is on the way.

What Should You Ask It First?

The best first prompt is the one that tells you what your pricing is doing in plain English. Most managers I talk to have rules they set a year ago and haven’t looked at since. Here are five prompts worth running in your first session. The first four come straight from the recipes in the docs, so you can paste them as written.

Read: “Explain what my pricing setup is actually doing, in plain English.” You get a one-sentence posture (“a premium hold with strong far-out protection but soft last-minute pricing”), a table of every active rule and what it does to your nightly rate, any rules that contradict each other, and the gaps: no floor, no ceiling, no weekend minimum.

Read: “Show me how my portfolio is pacing compared with this point last year.” Month by month, revenue, occupancy and ADR on the books now versus the same date a year ago, which months are furthest behind, and whether rate or occupancy is the problem in each.

Read: “Benchmark [listing] against last year and against its comps.” Your occupancy and ADR next to last year’s and next to what your comps are charging for the same dates, with the specific weekends where you are out of line, and the three dates worth acting on first.

Act: “Find revenue I’m leaving on the table in the next two weeks.” For one listing, it pulls the calendar, reservations and rules for the next 14 days, checks whether your comps are booked or also sitting open, and suggests prices and rule changes. It asks for your approval before saving anything.

Read: “Which open nights in the next two weeks should I lower the price on?” The question every host asks, answered against your comps and your own booking history rather than a hunch.

What I appreciate about these is how little you have to specify. You ask the way you’d ask a colleague, and the assistant knows which Adapt tools to use. I’m just telling Claude what to do at the end of the day.

Overview of AirDNA Adapt MCP prompts

Ten more prompts, sorted by the problem you are trying to fix

Once the first session is out of the way, these are the questions hosts and managers ask most. Each one maps to a specific set of Adapt tools, so the assistant goes straight to the right data instead of guessing. Anything that changes a rule or a strategy shows you a preview first.

Ten example prompts for the AirDNA Adapt MCP and the Adapt tools each one uses
PromptWhat it helps withAdapt tools it uses
1. “How much revenue will my Airbnb make over the next 3 months compared with last year?”Revenue and pacingget_portfolio_performance (pacing)
2. “Is my Airbnb priced right? Show me the suggested nightly prices for the next 60 days.”Pricingget_predicted_rates, preview_listing_rates
3. “How does my price compare with similar listings near me?”Pricing vs compsget_listing_comps, get_listing_comp_rates
4. “Should I offer a last-minute or weekly discount to fill my empty nights?”Discountssave_listing_rule, preview_strategy
5. “Why is my Airbnb occupancy low this month, and what should I change?”Occupancyget_listing_comp_performance, get_listing_market_context
6. “Set a revenue goal of $X for next year and adjust my pricing to hit it.”Revenue goalset_revenue_goal, apply_strategy
7. “What minimum stay should I set for weekends and holidays?”Minimum staysave_listing_rules
8. “Which price tier is my listing on, and what would my base rate be one tier up?”Price tierget_price_tier_options, set_price_tier
9. “Switch me to a more aggressive or more conservative pricing strategy and show me the difference first.”Pricing strategypreview_strategy, preview_custom_strategy, apply_strategy
10. “Which of my listings are pacing furthest behind last year, and which rules are holding them back?”Portfolio pacingget_portfolio_performance (pacing), get_listing_details (rules)

How I Run My Morning With It: The Daily Revenue Brief

The best revenue strategies are proactive, not reactive. Once the assistant can read everything in your account, you can have it build you a view of everything, organized around the problems you need to get ahead of. That is what I did.

A quick caveat first, because someone will mix this up: the dashboard I’m about to describe is something I built with Claude on top of the Adapt MCP. It is not a screen inside the Adapt app. It is a page Claude wrote for me, fed by the same tools you now have access to.

I called it the Daily Revenue Brief. It is the analysis tool I use to work out what a manager can change in their rules and pricing settings to earn more. Every morning it pulls reservations, a year of calendar in each direction, pricing rules and comp rates for every listing, and sorts them into the problems a revenue manager would otherwise have to dig for:

  • New bookings in the last 24 hours, and any unit that picked up two or more. Two bookings in a day usually means the price was under demand
  • Units pacing 20% or more behind last year on nights booked for the next 90 days, with last year’s numbers next to this year’s
  • Units with no new booking in the last 15 days that still have open nights in the next 60, with the likely cause: price vs comps, minimum stay, or open weekend nights
  • Units that are fully booked or blocked for the next 60 days. Selling out that far ahead usually means rates are under market
  • Bookings made further ahead than last year’s guests booked the same nights, at a lower rate. That points to far-out pricing set too low, and the owner will notice before you do
  • Open nights priced 20% or more above what the same night sold for last year
  • Split weekends and unbookable gaps created by a three-night minimum

AirDNA Adapt MCP Daily Revenue Brief

Each row comes with a suggestion: lower rates 9% on the open nights in the next 30 days, drop the minimum stay from three to two for check-ins inside 30 days, turn off the last-minute discount on a unit that is selling out. I don’t always agree with the suggestion. I always agree that I needed to look.

The small things make it usable at scale. I can mark a unit reviewed and it only comes back if things get worse. I can flag a row “talk to manager” and it collects on its own tab for the call. I can leave a note on a unit, have it draft the owner check-in email for me, and export the list for the owner relations team, who get as much out of it as I do.

Before this, most revenue managers pulled all of this from their PMS or their pricing tool and sifted through it. By the time they got to the problem, the problem had grown, and they were reacting instead of positioning. Now the problems find me, early enough to do something about them.

I’ve also built a few automations on top of the brief: a cap that keeps far-out nights from drifting more than 20% above what they booked for last year, and a routine that drops a weekend minimum to two nights when a checkout leaves a Saturday stranded. They run on a schedule and show me the plan before anything changes.

If you’re thinking “I don’t use Claude like that,” you don’t have to build any of this yourself. Once your listings are connected, reach out and we’ll help you set up a brief like this one, with the boundaries that fit your portfolio. You set the rules; it does the reading.

TALK TO OUR TEAM

Want Help Setting Up a Brief Like This One?

Tell us about your portfolio and our revenue management team will help you connect the Adapt MCP, set the boundaries that fit your listings, and build the checks you want running every morning.

Does It Change Prices Without Asking?

No. This is the question every manager asks first, so here is exactly how it works.

Reading is open; the assistant can look at anything in your account. Writing has layers. Adding or editing a rule is a real change, and it shows up in Adapt exactly as if you had added it yourself, so you can see it and undo it. Anything bigger has to be previewed first. Applying a strategy or pushing rates to your channel requires the assistant to show you the preview, and the preview generates a single-use confirmation that expires in ten minutes. If it didn’t preview it, it can’t push it. If it previewed one thing and tried to push another, it can’t do that either.

One warning from the help center worth repeating: if you ask the assistant to delete a season or an event, the pricing rules attached to it are deleted too, and that can’t be undone. Read the preview.

The concern I hear most from revenue managers about any automation is “I’m not in control.” I stand behind this setup 100% on that point. You have more control, because you can see what your pricing is doing in one question, and nothing changes on your channels until you say so. The tool processes the data. The judgment is still yours.

Adapt MCP Push confirmation

How Does Adapt MCP Compare With Other Short-Term Rental MCP Servers?

Several tools in this space have shipped MCP servers this year. They are not all for the same job.

Short-term rental MCP servers compared: what each reads, what it can change, the data behind it and availability
ServerWhat it readsWhat it can changeData behind itAvailability
AirDNA Adapt MCPYour listings, calendar, reservations, rules, comps and comp rates, pacing, market contextPricing rules, strategies, seasons, price tier, revenue goal; rate pushes with preview and confirmationAirDNA market and comp data, 15M+ active listingsBeta, public docs, Claude directory listing
PriceLabs AI ConnectorListings, performance, market trendsBase, min and max prices, customizations, nudgesPriceLabs market dataOpen beta, complimentary for a limited time
Beyond MCPPortfolio performance, pricing recommendations, market signalsAct on recommendationsBeyond dataPrivate beta, waitlist, included in plans
Wheelhouse MCPPricing preferences, rules, reservations, compsRules and custom ratesWheelhouse dataBeta
Hospitable MCP (PMS)Properties, calendar, reservations, messages, payoutsPricing updates, messaging, tasksYour PMS recordsLive, Host plan and up
Community “Airbnb MCP”Public Airbnb search results and listing pagesNothingScraped public pagesOpen source, guest search only

Watch the Walkthrough

On Thursday, November 12 I ran a one-hour session: a quick tour of what Adapt does, then the MCP, then the Daily Revenue Brief, including the owner check-in notes and the emails it drafts for you, live on real listings. Watch the recording here.

The right revenue strategy has a significant impact on your bottom line, and the biggest cost in most operations is the time between a problem appearing and someone noticing it. Connect your listings to Adapt, add the AirDNA connector in Claude, and ask it what your pricing is doing. Then watch the walkthrough and I’ll show you the rest.

¹ Based on AirDNA analysis of US entire-home listings with full-time availability, Aug 2025 to Jul 2026, matched on market, size, host scale and amenities. Observational comparison of comparable listings, not a guaranteed result.

Frequently Asked Questions

What is MCP?

MCP, or Model Context Protocol, is an open standard that lets AI assistants connect to other software and use its tools. When you add the AirDNA connector to Claude, Claude can read your Adapt data and take actions in Adapt the same way you would in the app. The technical docs are at mcp.airdna.co/docs.

Is there an official Airbnb MCP server?

No. Airbnb doesn’t offer an MCP server or a host API. The “Airbnb MCP” projects online are community search tools for travelers. The AirDNA Adapt MCP works on the listings you’ve connected to Adapt, using your own login.

What do I need to use the Adapt MCP?

An AirDNA login, a Claude or ChatGPT account, and at least one listing connected to Adapt. Once your listings are connected, the listing, pricing, comp and portfolio tools light up in your assistant. Until then it will report that it can’t find any properties.

Does it work with ChatGPT or Claude Code?

Yes. ChatGPT connects as a custom MCP server (the steps are in the connect section above), and Claude Code and Codex each take one command. Support for Cursor, VS Code and other assistants is being added.

What data does it not see?

It only sees your own listings. It never returns guest names or contact details, your listing’s street address, or other hosts’ listing records. Comp properties are described by title and structure, not by ID.

Can it make a change I didn’t ask for?

No. Rule changes are visible in Adapt immediately and can be undone. Strategy changes and rate pushes require a preview and your explicit approval, and the approval expires after ten minutes. Everything the assistant lists on a dashboard is a suggestion until you act on it.

ARTICLE SUMMARY

AirDNA revenue manager Jessica Spinks on the Adapt MCP, the Airbnb MCP server built for hosts: what Claude can read and change in your pricing, how to connect it in five minutes, and the daily checks and two automations she runs on it.

Topics:

AirDNA Product Features
Jessica Spinks

Jessica Spinks, CRME

AirDNA Founding Revenue Manager

Jessica Spinks, CRME, is Founding Revenue Manager at AirDNA, where she works directly with property managers and hosts on AirDNA Adapt, the company's dynamic pricing and revenue management tool. She spends her days inside real portfolios, from single listings to hundreds of units, helping operators pick a strategy, build comp sets, and catch the pricing problems that cost them bookings. She writes about how to optimize revenue management for short term rentals with AirDNA and how to streamline those processes with automated workflows.

Related

OwnerRez Integration: How to Set Up Dynamic Pricing With AirDNA Adapt

OwnerRez Integration: How to Set Up Dynamic Pricing With AirDNA Adapt

AirDNA Product Features
Blog

October 22, 2026

Get started with AirDNA today.

Powerful, easy-to-use tools that help hosts and investors succeed in the short-term rental market.

Try for free