AI for real estate investors: what the tools do, and the deal record none of them keeps

A deal is a hundred small conversations: the broker's hint about the seller's timing, the lender's phone call about the rate lock, the roofer's number from the walkthrough, the tenant who said the boiler was already loud in March. This guide names the AI tools investors actually run into, with what each one reads and what each one costs, as its maker published it, then turns to the record none of them keeps.

Short answer: the AI an investor meets comes in three kinds. Analysis tools work the numbers you give them. Data tools find properties and owners. The general assistants (ChatGPT, Claude, Gemini) read whatever you paste or upload and write a first draft of anything. None of them is a substitute for your own underwriting, and none of them knows what the lender said on Tuesday's call unless you put it somewhere.

That last gap is the expensive one. The spreadsheet is the easy part of a deal; you have done it fifty times. What goes missing is the correspondence: which version of the LOI the seller's broker actually has, whether the lender's "we can probably hold that rate" was before or after the appraisal came back, what the property manager promised the tenant in unit 4, and the contractor's verbal number that never made it into an email. Below: the tools, dated and priced from their own pages; how investors are using ChatGPT on deals; what never to paste; and a way to keep the deal record that works whatever else you run.

This page is not investment advice and makes no claim about any return, yield or valuation. It describes tools; the decisions stay yours.

AI tools for real estate investors: the names, grouped by job, with prices where published

These are tools an investor is likely to come across, grouped by the job each does. This is not a ranking; they do different jobs, and which fits depends on whether you buy single-family rentals, run a small multifamily portfolio or source commercial deals. Each description comes from the maker's own site and each price is as the maker publishes it, all read on 27 October 2026 from the US pages. Prices may exclude tax and change often, so confirm at the source before you buy. Where a maker publishes no price, we say so.

Deal analysis and underwriting

• DealCheck. A property analysis app: "Analyze any investment property in seconds." Its pricing page does not mention AI; it is on investor tool lists as a calculator, and the Left Main REI list of 18 October 2026 describes it the same way. Price: Starter at $0 ("No credit card required", limited to 15 saved properties), Plus at $10 a month and Pro at $20 a month, each paid plan with a 14-day free trial; yearly billing is offered with "3 months free!".
• Mashvisor. Rental analytics: "Invest Smarter with Industry-Leading Analytics." Its pricing page names an "AI Assistant" in the dashboard and a "Smart Property Finder" that, in its words, "organizes available properties in a decreasing order of investment opportunity score". Price: investor plans billed annually at $49.99 a month (Lite), $74.99 (Standard) and $99.99 (Professional); billed quarterly the Standard and Professional rates are higher ($99.99 and $119.99 a month).
• HouseCanary. Describes itself as an "AI-powered data & analytics platform for residential real estate" serving, among others, the "nation's top residential real estate investors, mortgage lenders, investment banks". Its CanaryAI is a "First-of-its-kind generative AI real estate assistant", and every tier lists "Full access to CanaryAI". Price: Basic at $19 a month or $190 a year, Pro at $79 a month or $790 a year, Teams at $199 a month or $1,990 a year; Enterprise by contact.

Finding deals and owners

• PropStream. "The Most Trusted All-in-one Solution That Provides Comprehensive Real Estate Data Nationwide." Every plan lists a "PropStream Intelligence Assistant", "AI Search", "Automated Valuation Modeling (AVM)" and "Property Condition Analysis (Photo AI)"; "AI Calling Scripts with Click-to-Dial" is an add-on on the entry plan and included above it. Price: Essentials at $99 a month or $81 a month billed annually, Pro at $199 or $165, Elite at $699 or $583, each with a "7-Day Free Trial With 50 Free Leads".
• DealMachine. "150 million properties. 240 million people. One search bar." Its AI Search page: "Type what you are looking for the same way you would explain it to an assistant", and "Every result comes enriched with owner contact info, mortgage details, equity estimates, and more." Price: the pricing page prices Basic and Pro by a records slider and says "Plans start at $99/mo · Cancel anytime"; Scale is $599 a month per package of 100,000 records.

Commercial data and deal management

• Reonomy. "The commercial property data platform for CRE professionals and commercial service providers", "An Altus Group Business". It names "AI & predictive analytics" and a "likelihood to sell" score that ranks properties by how likely they are to transact within 12 months. Price: standard plans "starting at $400/month on annual subscription"; bulk data and API by contact with sales.
• Dealpath. "The AI-Powered Operating System for Real Estate Investing", built for investment firms and their teams. Its home page describes "Dealpath AI" as a way to "Get answers about your deals and pipeline from AI tools", lists AI-powered data extraction under its Market Tracking feature and AI-recommended comps under Dealpath Connect. Price: not published; the call to action is "Request Demo".

Property-management AI

Leasing agents, maintenance triage and tenant-communication tools are a category of their own, and they matter more to the manager running the building than to the owner. They are covered in AI for property managers.

The general assistants: what each one reads

• ChatGPT (OpenAI). Reads what you paste, upload or, on some plans, connect. Its memory can draw on past chats and other sources you add; it is per account and vendor-scoped, and the summary it keeps is not a list you can audit line by line. The Left Main REI list of 18 October 2026 calls it "A common alternative for the same work" as Claude. Price: ChatGPT Plus is $20 a month, billed monthly, per OpenAI's help centre article "What is ChatGPT Plus?" ("Price: $20/month (billed monthly)", read 27 October 2026); the other plans are on OpenAI's pricing page.
• Claude (Anthropic). Reads what you paste or upload. The Left Main REI list says it "can read Word, PDF, Excel, PowerPoint, CSV and image files" and "search the web", and puts the general assistants to work "drafting seller scripts, summarizing documents, cleaning spreadsheets" and "pressure-testing a deal memo". Price: Anthropic's pricing page lists Pro at "$20 if billed monthly" or "$17 Per month with annual subscription discount ($200 billed up front)", Max "From $100 Per month", and Team standard seats at "$25 if billed monthly" or "$20 Per seat / month if billed annually".
• Gemini (Google). On Google's subscriptions page, Google AI Plus lists "Gemini in Gmail, Vids, and more"; Google AI Pro and Ultra list "Gemini in Gmail, Docs, Vids, and more", so from Pro up it reaches your Google documents as well as your mail. Price: Google AI Plus at $4.99 a month, Google AI Pro at $19.99 a month, Google AI Ultra "Starting at: $99.99/ month".

As of 27 October 2026: what each publishes

Equerry, the publisher of this guide, is not in the table on purpose: it does none of these jobs. It is a personal assistant that keeps the record of what each party in a deal told you and answers from it, and it is covered at the end of this page so the list above stays about other makers' products.

How investors use ChatGPT to run the business, as the people doing it describe it

The "how I use ChatGPT to run my real estate business" write-ups on page one of the search results have a common shape. They are practices, not products, and they are worth describing plainly, without endorsing any figure in them.

• Deal screening from pasted inputs. Andrew Jeffery, a multifamily investor writing the CRE Deal Junkie newsletter, described his method on 4 October 2024: tell the assistant your strategy and market, then for each property paste the rent roll (extracted from the offering memorandum and checked by hand) and a list of assumptions (rent growth, vacancy, debt terms, taxes, insurance, repairs, management, sale costs, exit cap rate, hold period, target return), ask for a maximum purchase price, then stress-test it: what rent would make a studio-to-one-bed conversion pay, how the answer moves at three exit cap rates. His own caveats: "They are not replacements for Excel, Argus, etc." and "You can't just upload an OM and say 'how much is this property really worth' and think the system is going to give you a valuable result." The work is in supplying the assumptions.
• A custom GPT loaded with your criteria. A guide published by theaiconsultingnetwork.com on 11 April 2026 walks through building a custom GPT for commercial deal analysis: load a market data summary, your underwriting template, two or three past deal analyses and your written investment criteria, then upload each new offering memorandum, rent roll and financials and ask it to extract the figures, show the math, list the top risks and flag missing data. Its own line on the limit: "a screening tool, not a replacement for due diligence."
• Drafting and document work. The Left Main REI list of 18 October 2026 puts Claude and ChatGPT to work "drafting seller scripts, summarizing documents, cleaning spreadsheets" and "pressure-testing a deal memo": in practice the LP update, the summary of a 70-page lease, the first draft of a letter to a seller.

Three things hold across all of them. The assistant knows only the figures you typed, so a wrong vacancy assumption produces a confident wrong answer. It has no knowledge of the building, the block or the seller's situation. And nothing it returns is advice; it is a structured first pass that you then underwrite properly. The practitioners say so themselves, which is why their write-ups are more useful than the tool lists.

What never to paste, and where each vendor's data terms are explained

Lender term sheets, tenant applications and files, your partners' financials and anything under an NDA from a seller's broker: before any of that goes into a chat window, read the document's own confidentiality terms and your operating agreement, and treat the question as one for your counsel, not for this page. Then check the tool. Anthropic's privacy page (read 27 October 2026) says Claude uses your chats to improve its models if "You choose to allow us to use your chats and coding sessions to improve Claude", a toggle called "Model Improvement" under Privacy Settings, and notes that incognito chats are excluded even when it is on. For ChatGPT, OpenAI's training setting, Temporary Chat and the business-plan defaults are laid out in Is ChatGPT confidential? and ChatGPT privacy: what OpenAI publishes. Whatever the setting, share what the task needs: a question about a loan's structure rarely needs the borrower's name or the address.

The deal and property record across brokers, lenders, contractors and tenants

Here is the part no tool above is built for. Say you have four properties and one deal in contract. In one week: the seller's broker emails a revised LOI, then calls to say the seller wants a shorter inspection period; your lender replies to a thread from a fortnight ago with a new rate-lock date; the property manager at the duplex forwards a tenant's complaint and says she "told him we'd look at it Friday"; a roofer gives you a number on the walkthrough, out loud, in the rain; your partner asks what the appraisal came in at. By the following Tuesday every one of those is in a different place: three email threads, a call you remember only roughly, a text, your head.

A chat assistant can remember more than it used to, but it keeps an auto-written summary of your chats, not a list you can audit, it is scoped to one vendor's account, and reaching your email that way means connecting a whole account rather than the one deal thread you care about. And the best information in a deal is said out loud, on a site or on a call. It reaches no tool until you record it.

A habit that works with any tool

An investor's record has a shape of its own: it runs per building, not per client, and it has to survive the close. The lender thread goes quiet for five years and then matters again at refinance; the roofer's walkthrough number is the baseline for the capex argument three winters later. So the habit is less about this week's deal than about the file you will reopen.

• One thread per deal and per building. Keep the broker, lender and attorney on a single running thread where you can; one thread is reconstructable, nine are not. When the deal closes, the thread becomes the building's file.
• Sixty seconds after every site visit and every call. Three lines in their words: what was said, what was promised, what you owe. Say the address and the person's name so you can find it later.
• Name the version. "LOI v3, sent 14 Oct, 21-day inspection" in the subject line or the first line of your note. Half the arguments in a deal are about which version someone is reading.
• Write down the dates that bite. Inspection period end, rate-lock expiry, estoppel due date, the day the tenant's renewal notice is owed. Missing one costs more than any tool on this page.
• Reread before you reply. Two minutes before the lender call, read what they said last time. Counterparties notice when you remember.

Or hand the file to a personal assistant

Equerry is a personal assistant on your iPhone that holds that record for you, from the things you send it. It works the way the assistant at a larger shop would: you copy it on the deal thread, you tell it what was said on the site, and you ask it later, this month or at refinance. The model stays yours; the correspondence around it stops living in your head.

• CC it on the deal thread. Your Equerry has its own email address. Add it to the LOI thread with the seller's broker and, for as long as it stays on the CC line, later replies keep arriving. Or forward just the one email: the lender's rate-lock message, the appraisal, the manager's note about unit 4. It never asks for your inbox, and the original stays in your own mailbox.
• Debrief by voice after the walkthrough. Record a voice note in the Equerry app in the car: the roofer's number, the seller's remark about timing, the smell in the basement. It is transcribed, so you can ask about it later.
• Send it the documents. Share the inspection report or the rent roll PDF from another app with the share sheet, or save a photo of the handwritten numbers from the site visit.
• Ask in plain words. "What did the lender say about the rate lock on Elm Street?" "Which version of the LOI went to the seller's broker, and what inspection period did it have?" "What did the manager promise the tenant in unit 4?" When an answer draws on a note you saved, it carries a source chip: tap it to see the passage Equerry used and open the original. What Equerry remembered from it is listed in Settings → Your memory, each line marked as from that email, and you can delete any line.
• Set a Reminder for the dates that matter. A Reminder is a push on a date you pick, with the text you wrote: the last day of the inspection period, the rate-lock expiry, the day the estoppel letters are due back.

Honest notes. Equerry does no underwriting, valuation, comps or deal sourcing; it will not tell you what a building is worth or whether to buy it. It connects to no MLS, CRM or property-management system, so the rent roll it knows is the one you sent, not the live one. It gives no investment advice, and it does not send email, make calls or sign anything for you. It knows what you send it, plus your calendar if you allow it, and nothing else. A CC is visible to everyone on the thread, so if the seller's broker seeing an extra address is awkward, or your partners would ask about it, forward the email instead. And sending something is still sharing it: what you send is processed to make it searchable and answerable, and reaches the outside providers named on our AI disclosure page. The confidentiality question above applies here too: a lender's term sheet under restriction is restricted wherever you put it, and tenant files belong with whoever is bound to protect them.

If the correspondence is the problem you came here with, the next reads are how to keep track of important emails and how to keep track of follow-ups. If you manage buildings for owners rather than own them, AI for property managers is the closer page; if you sell them, AI for real estate agents.

Last updated 2026-10-28.

All product names, logos, and brands are property of their respective owners. Equerry is not affiliated with, endorsed by, or sponsored by any other product mentioned; comparisons are provided for informational purposes.