Claude + Airbnb = $15,000/Month

A guy I know clears about $15,000 a month from Airbnb. He does not own a single apartment.

Most people think Airbnb means a mortgage or a paid-off second home. It does not. The model is called rental arbitrage: you lease a place long-term, you list it short-term, and the spread is your profit.

It has been around for a decade. The only thing that changed in 2026 is that the part that used to take weeks now takes twenty minutes, because Claude does the analysis.

Here is the whole system.

The gap you are selling

You sign a long-term lease. You furnish the place. You list it nightly on Airbnb. The difference between the rent you pay and the revenue you collect is yours.

No buying. No equity. Just a spread, run across as many units as you can manage.

The catch is finding markets where the spread is actually big enough, and that is exactly the boring, slow research Claude is good at.

The math on one unit

Illustrative numbers, the kind you would verify for your own city before signing anything.

Say a one-bedroom rents long-term for $1,400 a month. Comparable Airbnb listings nearby run about $145 a night at roughly 74% occupancy.

1Gross  = $145 × 0.74 × 30            = $3,219 / month
2Rent                                 = -$1,400
3Cleaning (9 turnovers × $55)         = -$495
4Utilities + supplies                 = -$230
5Airbnb fee (~3%)                     = -$97
6Net per unit                         ≈ $997 / month

One apartment nets around a thousand. The number in the title is not one unit. It is a handful of them, or a few in higher-rate markets, run by one person with Claude doing the desk work.

Why this used to take weeks

The old workflow was Zillow, plus AirDNA, plus spreadsheets, plus reading every lease, plus checking city rules. Four to six hours per unit before you even called a landlord. Most people quit before their first lease.

Claude collapses that into a prompt. Drop in a city, get a verdict.

The prompts

These are yours to copy. Change the brackets and run.

1. Market analyzer.

1You are an Airbnb rental-arbitrage analyst. Legal methods only.
2
3Market: [CITY / NEIGHBORHOOD]
4Property: [1BR / 2BR / studio], long-term rent $[X]/month.
5
6Give me:
71. Typical Airbnb nightly rate (ADR) for comparable listings here
82. Average occupancy % (use AirDNA-style benchmarks; say if estimated)
93. Peak vs slow months
104. Net monthly profit per unit:
11   Net = (ADR × occupancy × 30) - rent - Airbnb fee(3%)
12         - cleaning - utilities - supplies. Show the math.
135. Verdict: STRONG (>$800 net) / MARGINAL ($400-800) / SKIP (<$400)
146. Red flags: oversupply, seasonal collapse, HOA or city STR limits
157. Note if short-term rental and subletting are even allowed here.
16
17Be specific, show every number, and flag anything you are guessing.

2. Pricing optimizer.

1You are an Airbnb pricing strategist.
2
3Listing: [NEIGHBORHOOD], [N] bedrooms, sleeps [N].
4Current price $[X]/night, occupancy [X]%.
5
6Build me:
71. A 12-month price calendar (peak / shoulder / low season)
82. Weekend premium (% over weekday)
93. Last-minute discount (days out, % off)
104. Event and holiday bumps for this area
115. Minimum-stay rules by season to cut cleaning cost
126. Projected RevPAN (ADR × occupancy) before vs after.
13
14Output as one clean table.

3. Guest ops.

1Draft my full Airbnb guest message stack for [PROPERTY, CITY]:
2- instant inquiry reply
3- booking confirmation + check-in instructions
4- house manual + wifi + rules
5- local recommendations (food, transit, one hidden gem)
6- checkout steps
7- review request
8
9Warm, short, skimmable. One message per section.

Flat pricing alone leaves 15 to 25% of revenue on the table, so prompt 2 usually pays for the whole month by itself.

The other lane: sell it as a service

You do not have to run units yourself. The same prompts optimize other people's listings, and most hosts will pay for it.

Enrich their photos, rewrite their titles, build their pricing rule, write their guest stack. Better photos alone lift bookings up to 40% with a higher nightly rate, so the before-and-after closes the sale on its own. Charge per listing or on a monthly retainer.

Either way, Claude is the operator. You are the one who shows up.

What actually goes wrong

The lease is the one thing you cannot skip. Not every landlord allows subletting, and some cities require a short-term-rental license. Read the contract, check the city, then sign. Never the other way around.

Seasonality is the quiet killer. Some markets lose more than half their occupancy in winter. Pull twelve months of data, not a three-month average.

Guests message at 2am. You either answer or you hire a co-host for 15 to 20% of revenue.

And new listings get a short algorithm boost, then you compete on photos, reviews, and price. The first ten reviews matter more than the next hundred.

Two things to do now

Bookmark this. In six months you will either thank yourself or wish you had started.

Then open Claude, drop your own city into prompt 1, and see if the math works where you live. Two minutes.

The strategy was never the hard part. The friction was, and the friction is gone.

I break down AI tools and the money being made with them like this regularly. Follow for the next one, and join my Telegram: https://t.me/+lzSPoQ0svRU1ZWZi