AirDNA does not ship an official MCP server, and its API is enterprise-only, provisioned by an account manager rather than a self-serve key. You can still connect AirDNA to Claude today: export market metrics as CSV files from a paid AirDNA dashboard, upload them to Claude, and run your comp sets, seasonality breakdowns, and revenue projections in plain English. Enterprise customers who already hold an API token can additionally wrap it in a small custom MCP server.
Your three routes, honestly ranked
Most guides in this series wire up a live MCP server against a self-serve API. AirDNA is different, so we checked every route before writing this page.
CSV export + Claude
Export Occupancy, Rates, and Revenue data from AirDNA on a Research plan or above, then upload to Claude for analysis. No code, no sales call.
Enterprise API + MCP
AirDNA's API token is set up at purchase by your account manager. If you have one, a custom MCP server gives Claude live queries.
Official AirDNA MCP
As of July 2026, AirDNA has released no official MCP server. Marketplace "AirDNA MCP" listings are third-party scrapers, not AirDNA.
What you can ask Claude once the data is in
These all work with nothing more than exported CSVs uploaded to a chat or a Claude Project. The examples assume an investor or operator lens because that is who uses AirDNA.
"Here are 36 months of occupancy and ADR for two markets I'm comparing. Which has the more stable shoulder season, and how big is the January trough in each?"
"From this export of my competitor list, which comps out-earn my listing on RevPAR, and what do they share: bedroom count, amenities, or rate strategy?"
"Build a 12-month revenue projection for a 3BR in this market using the exported ADR and occupancy data. Give me conservative, base, and aggressive cases."
"At a $485,000 purchase price with 20% down, what occupancy does this market's ADR history imply I need to break even each month? Show the math."
"Compare my monthly revenue export against the market's revenue curve. In which months am I leaving the most money on the table, in dollars?"
"I've uploaded exports for all five markets I operate in. Rank them by year-over-year RevPAR trend and flag any market where supply growth is outpacing demand."
Prerequisites
- A paid AirDNA subscription. According to AirDNA's help center, CSV export is not available on the free plan; it is included from the Research plan up. AirDNA's tiers are Free, Research, Host, and Advanced.
- Claude, any surface that accepts file uploads: claude.ai in the browser, the desktop app, or a Claude Project if you want the data to persist across chats.
- No code and no API key for the primary route. The MCP route further down needs an enterprise API contract and Node.js.
Route 1: CSV exports + Claude (works today)
Export the market data from AirDNA
Log in at app.airdna.co and open the market you are researching. According to AirDNA's help center, market-level metrics export from the Occupancy, Rates, and Revenue chart pages via the export button in the top right corner of each chart, and the files come down as CSV, ready for Excel or Google Sheets. Grab all three for a full picture. On the Research plan you get 36 months of history; the Advanced plan reaches back to 2017.
Two useful extras while you are in there: your custom competitor list (Research plan and up) exports the comp-level view, and any Rentalizer address report can be saved as a PDF, which Claude also reads.
Upload the files to Claude
Open a new chat at claude.ai (or the desktop app) and attach the CSVs with the paperclip. If this is recurring research, create a Project called something like "Market research: Broken Bow" and add the exports to its knowledge, so every future chat starts with the data loaded.
Then anchor the analysis with context Claude cannot guess:
Run the analysis
Start with one of the prompts from the section above, then push into follow-ups: charts, month tables, sensitivity ranges. Two habits make the output dramatically better:
- Make Claude show its work. "Walk me through the calculation" catches unit mistakes (monthly vs annual revenue, percentage vs decimal occupancy) before they compound into a bad projection.
- Keep projections tied to the export. Ask Claude to cite which rows and months drive each conclusion. If a number did not come from your file, it should not be in the model.
Re-export monthly
AirDNA data updates on an ongoing basis, and your uploaded snapshot does not. For an active market thesis, re-export and replace the Project files monthly, then ask Claude what changed: "Compare this month's export to the previous one. What moved in occupancy, ADR, and active supply?" That delta question is often more valuable than the original analysis.
Route 2: The enterprise API (if you have a token)
AirDNA does have a real API, but it is not self-serve. According to AirDNA's enterprise help center, an API access token is "set up at purchase by your account manager," and getting access starts with the contact form and an account executive. The same help center describes what the API covers: property valuations via Rentalizer, listing-level ADR, occupancy, and revenue data, market data and market search, forward property availability, and Smart Rates daily rate recommendations. Technical documentation lives at apidocs.airdna.co, and AirDNA's docs describe authentication as a Bearer token in the Authorization header.
If your company already holds a token, the same pattern from our PriceLabs and OwnerRez guides applies: a small Node.js MCP server that wraps the endpoints you are licensed for, registered in claude_desktop_config.json. The skeleton looks like this; fill in the endpoint paths from your own API documentation, since AirDNA's full reference sits behind the enterprise agreement and we will not guess at paths we cannot verify.
#!/usr/bin/env node
import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
import { z } from "zod";
const TOKEN = process.env.AIRDNA_API_TOKEN;
async function airdnaGet(url, params = {}) {
const u = new URL(url);
for (const [k, v] of Object.entries(params)) {
if (v !== undefined && v !== null) u.searchParams.append(k, String(v));
}
const res = await fetch(u, {
headers: {
Authorization: `Bearer ${TOKEN}`,
Accept: "application/json",
},
});
if (!res.ok) throw new Error(`AirDNA API ${res.status}: ${await res.text()}`);
return res.json();
}
const server = new McpServer({ name: "airdna", version: "1.0.0" });
// Wrap each endpoint you are licensed for. Take the exact paths and
// parameters from your enterprise docs at apidocs.airdna.co rather
// than guessing them. One read-only tool per endpoint, e.g.:
server.tool(
"market-metrics",
"Fetch market-level metrics for a market ID from your licensed AirDNA API package.",
{ market_id: z.string().describe("AirDNA market identifier") },
async ({ market_id }) => {
const data = await airdnaGet("YOUR_LICENSED_ENDPOINT_URL", { market_id });
return { content: [{ type: "text", text: JSON.stringify(data, null, 2) }] };
}
);
await server.connect(new StdioServerTransport());Troubleshooting
There is no export button on my AirDNA charts
Data export is not available on AirDNA's free subscription. According to AirDNA's help center, CSV export is included from the Research plan up. If you are on a paid plan and still do not see it, confirm you are on a chart page (Occupancy, Rates, or Revenue) rather than the market overview; the export control sits in the top right of the chart.
Claude misreads the columns or mixes up units
AirDNA exports do not explain themselves. Before analyzing, tell Claude what each file contains ("occupancy is a monthly percentage of booked nights over available nights") and ask it to echo back the columns and date range it sees. If a projection looks off by roughly 10x or 12x, suspect a monthly-vs-annual or percent-vs-decimal mixup and ask Claude to re-derive one month by hand.
My upload is too large or the chat slows down
Trim the export in a spreadsheet first: keep the metric columns and the date range you actually need. Separate uploads per market also work better than one merged mega-file, and a Claude Project keeps them organized without re-uploading every chat.
Can I just point Claude at app.airdna.co with a browser tool?
Technically possible with browser-automation MCP servers, but it is a bad trade: logged-in dashboards are slow to drive, brittle when the UI changes, and scraping may sit outside AirDNA's terms of service. The CSV export exists precisely so subscribers can take their licensed data with them. Use it.
What about the "AirDNA MCP" servers on Apify and similar marketplaces?
Those are third-party scrapers that estimate AirDNA-style metrics or enrich listings, not AirDNA products, and they are unaffiliated with AirDNA. If your workflow depends on AirDNA's actual numbers, use your subscription's export or the official enterprise API rather than a lookalike feed.
FAQ
Does AirDNA have an official MCP server?
No. As of July 2026, AirDNA has not released an official MCP (Model Context Protocol) server. Nothing appears in AirDNA's documentation, and the "AirDNA" entries on MCP marketplaces are third-party scrapers. If AirDNA ships one, this page will be updated; a data platform with an existing enterprise API is a natural candidate to do so.
Does AirDNA have a public API?
Not a self-serve one. According to AirDNA's enterprise help center, API access tokens are set up at purchase by your account manager, and the path in is the contact form. The technical docs live at apidocs.airdna.co.
Can I export data from AirDNA on a normal plan?
Yes, on paid plans. AirDNA's help center lists CSV export from the Research plan up, from the Occupancy, Rates, and Revenue chart pages, plus unlimited Rentalizer PDF reports. The free plan cannot export.
Can Claude really do useful analysis on AirDNA exports?
Yes. Claude accepts CSV uploads in chats and Projects and can compute, chart, and compare across files. Everything in the "what you can ask" section above runs on exported data alone. The quality driver is context: tell Claude what each column means and make it show its calculations.
Is "MarketMinder" the same thing as AirDNA?
MarketMinder is AirDNA's market research dashboard, the product you log into at app.airdna.co. AirDNA's own help center uses the name in its export documentation, so if you see "export from MarketMinder," that means the same Occupancy, Rates, and Revenue pages described in this guide.
Other guides in this series
This page is part of RapidEye's Claude guides for the STR stack: one guide per platform, each verified against what the vendor actually offers. If you are wiring several of these together, our API page covers what a fully connected inspection layer looks like.
Connect PriceLabs to Claude
A working read-only MCP server against the self-serve Customer API. The full MCP walkthrough lives here.
Connect OwnerRez to Claude
Bookings, properties, guests, and payments through a personal access token.
All Claude guides
PMS, pricing, screening, smart locks, and turnover ops. Every guide, one index.
Sources & references
- Where Can I Download Data in AirDNA's MarketMinder? (AirDNA Help Center) https://help.airdna.co/hc/en-us/articles/15164930853773-Where-Can-I-Download-Data-in-AirDNA-s-MarketMinder-
- AirDNA Subscription Plans (AirDNA Help Center) https://help.airdna.co/en/articles/8062197-airdna-subscription-plans
- How Can I Connect to AirDNA's API? (AirDNA Enterprise Help) https://enterprise-help.airdna.co/en/articles/8062183-how-can-i-connect-to-airdna-s-api
- Using AirDNA's Application Programming Interface (API) (AirDNA Enterprise Help) https://enterprise-help.airdna.co/en/articles/8185669-using-airdna-s-application-programming-interface-api
- AirDNA API documentation https://apidocs.airdna.co/
- AirDNA Enterprise API docs (Bearer authentication) https://docs.airdna.co/
- Model Context Protocol TypeScript SDK https://github.com/modelcontextprotocol/typescript-sdk