The best AI prompts for real estate ads, 20 to copy (2026)

A good prompt for real estate ads carries four details: the property, the area, the buyer profile and the tone. With those, ChatGPT, Claude, Copilot or Gemini write a useful description in minutes. This guide collects 20 prompts to copy. To apply them with a method, see the AI program for real estate agencies.

Add Founderz as a preferred source

What you’ll take from this

This practical guide is a free resource ready to use in your agency’s daily work. This is what it includes:

• 20 prompts to copy and paste: sale and rental descriptions, headlines, versions per portal, replies to leads and owner outreach.
• Which prompt to use in ChatGPT, Claude, Copilot and Gemini, based on what each tool does best with your portfolio.
• The method for writing your own prompts: role, property, buyer, format and what to avoid.
• The limits of AI in ads: unverified data, regulations and lead privacy.

Every week you write the same texts: remodeled apartment, three bedrooms, exterior, very bright. Each ad eats half an hour you could spend talking with clients, and in the end they all sound alike. If you paste “write an ad for an apartment” into an AI, you get a correct but interchangeable text, because it knows nothing about your property or who you want to sell it to.

A common case: a small agency with 40 listings drafts every description, one version per portal and template replies to inquiries in one morning. The time saved on writing goes to calling and showing apartments. No text goes out before an agent checks square footage, price and charges.

Real estate agents and real estate marketing teams that use AI to win clients look for ready-to-copy prompts that let them save time, automate repetitive tasks and write an attractive description of each property with its key features: finishes, pool, sea views, neighborhood or parking. In real estate, AI also helps prepare a market analysis or value a home with comparables, always with your review.

These prompts start from ChatGPT, but work the same in Claude, Copilot or Gemini. If you want an overview before you start, this guide to artificial intelligence in real estate puts you in the full context.

What data a prompt needs to write a good ad

An ad’s result depends on the data you put into the prompt. With vague information, AI returns generic text; with specific data, it writes tailored descriptions. These are the ones worth always including:

• Property type: apartment, penthouse, house, commercial space, parking spot.
• Location: neighborhood, proximity to transport, schools or services.
• Square footage and layout: built area, bedrooms and bathrooms.
• Transaction: sale or rental, with the price if you want to include it.
• Distinctive features: recent renovation, terrace, exposure, elevator, garage, sea views or pool.
• Buyer or tenant profile: family, investor, first-time buyer, young professional.

The last point makes the difference. The same apartment reads differently for a family looking for schools nearby than for an investor eyeing returns. When the prompt defines the profile, AI adjusts the tone and highlights what matters to that reader.

ChatGPT prompts for real estate ads

ChatGPT is the most versatile option for day-to-day work: it writes, rewrites, works with files (a portfolio spreadsheet, for example) and adapts well to the tone of each channel. Copy the prompts as they are: what is in brackets is what you change.

Sale and rental descriptions

Description of an apartment for sale

Act as a real estate copywriter specialized in [second-hand apartments in Miami]. Write the description of an apartment for sale with this real data: [property type], [neighborhood and nearby streets], [square footage], [bedrooms and bathrooms], [renovation, terrace, exposure, elevator], [price].
The target buyer is [a family with one child looking for schools nearby].
Format: a headline of at most 8 words and a 120-word body in two paragraphs. Highlight [the natural light] and close with a call to action to book a showing.
Professional and friendly tone, without exaggerating. Do not add data I have not given you. If something a buyer would want to know is missing, ask me before writing.

Description of a rental apartment

You are a rental real estate agent in [Chicago]. Write the ad for a rental apartment with this data: [neighborhood], [square footage], [bedrooms], [furnished or not], [monthly rent], [deposit and minimum lease length], [utilities included].
The target tenant is [a young professional who works downtown].
Format: headline of at most 8 words, 4 bullets with the essentials (rent, availability, terms, transport) and a 60-word paragraph.
Be clear about the terms to filter out uninterested inquiries. Do not use phrases like "unique opportunity".

Headlines, versions per portal and social media

Headlines to test CTR

This is the description of my ad: [paste the ad].
Give me 8 headlines of at most 70 characters for the portal [Zillow or Realtor.com]. Each must highlight a different benefit (location, light, renovation, price, terrace, transport, schools, returns).
Present the result in a table with the headline, the benefit it highlights and the character count. Avoid empty adjectives like "amazing" or "unique". Recommend the two I should test first and why.

One ad for a portal, Instagram and WhatsApp

Based on this property description: [paste your ad], generate three versions: 1) real estate portal, formal, 100 words; 2) Instagram, friendly, with 3 emojis, 4 local hashtags and a call to action; 3) a 30-word WhatsApp message to send to interested contacts.
Keep the exact data (square footage, bedrooms, price) identical in all three and change only tone and length. Present them in a table with one column per channel.

Script for a 30-second reel

You are a content creator for a real estate agency in [Austin]. Write the script of a 30-second reel to show [a penthouse with a terrace downtown], aimed at [young couples who want to buy their first home].
Structure: a 3-second hook, 4 shots (say what to film in each one and what the on-screen text says) and a final call to action: [message me to book a showing].
Natural tone, short sentences. Do not promise a price or returns I have not given you.

Claude prompts for real estate ads

Claude stands out when you work with long documents and want a careful draft: you can upload a development brochure, an appraisal report or your agency’s style guide and ask it to write following them.

Ads from long documents

Ad from a new development brochure

I am attaching the sales brochure of a new development as a PDF. Write the portal ad aimed at [families buying their first home].
Use only data that appears in the document and cite in parentheses the page each fact comes from (floor areas, finishes, delivery dates, common areas).
Translate the technical finishes into benefits for the buyer. Format: headline, 120-word paragraph, 6 bullets and a table with unit types and floor areas.
Mark with (unverified) any doubtful or contradictory data in the brochure.

Ad in your agency’s voice

I am attaching my agency's style guide [file] and three ads that have worked well [paste or attach].
First summarize in 5 rules how my agency writes. Then write the ad for [property and data] following those rules.
At the end tell me which tone decisions you made and which sentence you would hesitate to publish.

Review and delicate replies

Audit of your listing portfolio

I am attaching [25] ads from my portfolio in one document.
Review: 1) tone inconsistencies between ads; 2) data that contradicts itself (square footage, year, fees); 3) ads under 60 words; 4) phrases repeated in more than three ads; 5) claims that need checking, such as "best area" or "guaranteed returns".
Return a table with the ad, the problem and the proposed fix, ordered from most to least urgent.

Reply to an owner who wants a high price

An owner wrote to me: [paste the message]. They want to list their apartment at [price] and my comparables analysis points to [price range], with this supporting data: [3 recent sales in the area with price and square footage].
Write a reply of at most 150 words: acknowledge their expectation without agreeing, explain the range with the data and propose [listing at a launch price and reviewing after 4 weeks].
Calm and respectful tone, no jargon. Do not promise sale timelines.

Copilot prompts for real estate ads

Copilot is useful if your agency already works in Microsoft 365: it can read your Excel and Word files and draft in Outlook or Teams without leaving the app, depending on your license. Name the file in the prompt and check your company policy before using client data.

Descriptions from your portfolio in Excel

Using the table in this Excel sheet [sheet name], add a "Description" column with a 70-word text per property.
Use only the columns [type, neighborhood, square footage, bedrooms, features, buyer profile]. [Friendly and professional] tone. Do not repeat the same opening sentence in two consecutive rows.
Add another column with the word count and a third with "REVIEW" if a key data point is missing in the row.

Reply to an inquiry in Outlook

Act as a real estate agent. A client interested in [property] wrote this email: [paste the inquiry or use the selected email]. Their main objection seems to be [price / condition / location].
Write a reply that addresses the concern, provides a concrete fact from [the property sheet] and proposes two time slots for a showing.
Maximum 120 words, friendly tone, no pressure. Also give me a subject line under 50 characters.

Property report in Word for the owner

Using the information in this Word document [document name] and the comparables sheet [file name], create a one-page report to present to an owner.
Include: property summary, 3 comparables with price and square footage, recommended price range and a marketing plan (portals, social media, showings).
Format: short headings, a table for the comparables and a maximum of 350 words. Do not invent any figure that is not in the files.

Follow-up after an owner meeting in Teams

From the transcript of this Teams meeting with an owner, summarize in 5 points what we agreed and what is still pending.
Then write a 100-word follow-up email with the next steps, who does what and a proposed date for [the photo shoot].
If any agreement is unclear in the transcript, mark it "CONFIRM" instead of assuming it.

Gemini prompts for real estate ads

Gemini fits if your agency lives in Google Workspace: Docs, Sheets and Gmail. These three prompts are built for that workflow.

Property sheet in Google Docs with FAQ

Write in this document the sheet for [a 4-bedroom townhouse] for [families relocating for work].
Structure: 120-word description, 6 bullets with the essentials and 4 FAQs with one- or two-sentence answers (HOA fees, nearby schools, parking spaces, move-in date).
Use only this data: [paste the data]. If an answer is not in the data, write "[ASK THE OWNER]".

Ads per portal from a spreadsheet

These are my properties in a spreadsheet: [paste the rows with type, neighborhood, square footage, bedrooms, price and main feature].
For each row write a headline of at most 70 characters and a 90-word description in the format of [portal].
Return it in columns ready to paste: headline, description, headline character count. Do not use shouting capitals or exclamation marks.

Follow-up email after a showing in Gmail

Write an email to [Marta], who visited [a 3-bedroom apartment in Wicker Park] yesterday and mentioned a concern about [street noise].
Body of 90 words: thank her for the visit, answer the concern with this real fact: [double-glazed windows and bedrooms facing the inner courtyard] and propose [a second visit in the afternoon].
Subject line under 45 characters. Warm tone, a single call to action.

More AI prompts to win owners and leads

Winning listings and following up leads is where the time you recover shows most. These prompts work in any of the four tools.

Outreach letter to owners in an area

You are a real estate agent in [East Austin]. Write a 130-word letter to owners of apartments in [neighborhood] who do not have their home listed.
Include: why you are writing to them (buyers are searching in the area, without exaggerating), a real fact: [I sold 6 homes in the neighborhood this year], and a concrete offer: [a free, no-obligation valuation].
Respectful tone, no pressure. Avoid "last chance" and any price promise.

Script for the first call with a lead

Prepare a script for the first call with a lead who inquired about [property] on [portal].
Include: a 15-second opening, 5 qualifying questions (budget, timeline, financing, who decides, what they really want), how to answer "it is too expensive" and "I need to think about it", and a close that proposes a showing.
Format: numbered list, short sentences I can say out loud.

Comparison of two properties for an undecided buyer

My client is torn between these two properties: A) [data of apartment A] and B) [data of apartment B]. Their priority is [proximity to work and a budget of $300,000].
Make a comparison table with price, square footage, price per square foot, fees, transport and condition. Then write a neutral 80-word paragraph that helps them decide.
Do not favor one over the other if the data does not justify it, and point out what data the client should check on each.

Listing win post for LinkedIn or Facebook

Write a post for [LinkedIn] about [a home sold in 28 days] with these real facts: [asking price, number of showings, what price adjustment was made].
Tell the process in 120 words: the challenge, what we did and the result. End with a question for owners who are thinking about selling.
Professional and friendly tone. Do not use buyer data or the exact address.

How to fit the prompts into your agency’s workflow

AI performs better when you build it into your process than when you use it sporadically. Place the prompts in four stages:

1. Capture: you gather the property data into a fixed template that you then paste into the prompt.
2. Writing: you generate the ad and its variations for portal, social media and email.
3. Publishing: you review, correct data and publish on each portal.
4. Follow-up: you use prompts to answer inquiries and objections as leads come in.

You can go a step further by connecting AI with your CRM and WhatsApp: when an inquiry comes in, the CRM logs the lead and you generate a first reply with a prompt in seconds. Some agencies use chatbots for first contact and step in when the lead is hot. Automation sustains the volume, but judgment stays yours: AI prepares the draft and you decide what goes out. The same logic works for launching new development promotions. If you are looking for which tools fit your agency, this overview of the best AI tools for real estate agencies helps you choose.

How I built these prompts for real estate ads

All the prompts above come from the same scheme. A prompt is not a loose question: it is a crafted instruction, with role, context and output format. These are the fields I always use, in this order:

Five tips for writing your own prompts

• Give real data. AI fills the gaps with inventions. Paste the exact features and say not to add any.
• Define the buyer. “For a family” and “for an investor” produce two different ads for the same apartment.
• Limit the format. Words, bullets and characters prevent long, empty texts.
• Let it ask. Add “if any data is missing, ask me before writing” and you will see what information you were missing.
• Save what works. When a prompt serves you, keep it in a shared document and change only what is in brackets.

From a weak prompt to a strong one

Weak prompt:

Write an ad for an apartment.

The same prompt, with the fields:

You are a real estate copywriter. Write the ad for a remodeled 915 sq ft, 3-bedroom apartment in [Wicker Park], with a terrace and an elevator, aimed at a family looking for schools nearby.
Format: 8-word headline and 120-word body. Professional and friendly tone.
Do not add data I have not given you. If any is missing, ask me.

The first produces a text that would fit any apartment. The second produces an ad with an angle, a buyer and a voice. That difference in context steers the model, although it does not guarantee the same result every time. If you would rather not write it by hand, the prompt generator above assembles these fields for you.

How the prompt changes by AI tool

The scheme is the same, but each tool performs better with a certain type of task and file.

ChatGPT

The most flexible for free-form writing, tone adjustment and files. Paste or attach your portfolio and ask for the result in a table.

Gemini

It integrates with Google Workspace (Docs, Sheets, Gmail). Name the document and the columns you want it to work with.

Copilot

It works inside Word, Excel, Outlook and Teams with your files, depending on your license. Name the file or sheet in the prompt.

Claude

Upload long documents (development brochure, appraisal report, style guide) and ask it to cite where each fact comes from.

Tools change plans often: check the current terms before you subscribe.

AI limits in real estate ads: where human judgment rules

AI does not verify data: it reproduces what you give it. It does not check whether the square footage is correct, whether the price is current or whether a claim meets regulations. That verification is yours.

• It does not verify facts. Square footage, price, year built or charges: everything must be checked before publishing.
• It tends toward generic text. With a weak prompt, it returns interchangeable descriptions that do not sell the property.
• It does not know the legal context. In Spain, for example, the ad must state the property’s energy rating, and AI does not add it on its own.

FAQ about AI prompts for real estate

These are the most common questions from people who start using AI to write ads and handle leads at their agency.

How is artificial intelligence used in real estate marketing?

It is used to write ad descriptions and headlines, adapt texts to each portal, create social media content and answer client inquiries. It works with prompts: you provide the property data and the tool generates a draft that you review and publish.

How do I write an apartment description with ChatGPT?

Write a prompt with a role (“act as a real estate copywriter”), the real property data, the buyer profile, the format (headline plus 120 words) and the tone. Then review square footage, price and features, because AI does not verify them.

Which ChatGPT prompts work for a real estate agency?

The most used are the sale or rental description, headlines to test clicks, a version per portal and social media, replies to inquiries and an owner outreach letter. They are all above, ready to copy.

Can I use Claude for real estate ads?

Yes. Claude works well with long documents: upload a development brochure or your style guide and ask it to write using only that data and to cite the page of each fact.

How do I use Copilot to write real estate ads?

Use it inside Excel to generate one description per row of your portfolio, in Outlook to answer inquiries and in Word for the owner report. It depends on your Microsoft 365 license.

Which prompt do I use in Gemini for my agency?

If you work in Google Workspace, ask it to write the property sheet in Docs, generate the ads per portal from a spreadsheet or draft the follow-up email in Gmail. The prompts are in its section.

How do I get leads in real estate with AI?

AI does not capture leads on its own: it speeds up content production and follow-up. Use prompts for clear ads per portal, owner outreach letters and fast replies to every inquiry, and spend the time saved calling and showing homes.

How do I know whether ChatGPT recommends my agency when someone asks about properties?

Ask it the way a buyer in your area would, for example “which real estate agency do you recommend in [city]”, and note whether it names you and what it says. Repeat with several questions and compare month to month; answers can change between queries.

Is it safe to use leads’ personal data in AI tools?

It depends on how you do it. Enter only the minimum data, avoid full names, phone numbers or addresses when not needed and check the tool’s privacy policy.

Your next step with AI prompts for real estate ads

Back to the problem at the start: ads that all sound alike and half an hour lost on each. With a prompt that has a role, real data, a buyer and a format, those flat texts become ads that speak to whoever reads them. That judgment is built by practicing with real cases.

If you want your team to apply this with a method, the AI training for real estate agencies at Founderz is designed to bring artificial intelligence into real capture, writing and follow-up tasks, with a practical and responsible approach.