Home Search Moved Into ChatGPT. What the Zillow and Realtor.com Apps Mean for Agents

By The Professions Cloud Editors Updated October 2026 · 6 sources · 3 min read

Prices and facts checked October 13, 2026. We don't accept payment for rankings.

What happened: on October 6, 2025, at OpenAI's Dev Day, Zillow became the first real estate company with an app inside ChatGPT (RISMedia). In April 2026, Realtor.com launched its own ChatGPT app, focused on the "pre-search" stage of affordability and neighborhoods (Online Marketplaces).

Why it matters today: a growing share of buyers start with a chat, not a search box. They ask an AI what they can afford and which neighborhoods fit, then click through to a portal. Where those clicks land decides who gets the lead.

How the two apps work

Sources: Zillow; HousingWire; Online Marketplaces.

How buyers actually use it

Buyers don't type filters into a chat. They describe a situation, and the AI turns it into a search. Typical prompts look like:

• "Zillow, show me 3-bedroom homes under $600,000 in Denver with a big backyard"
• "What can I afford on a $120,000 salary with $40,000 down?"
• "Compare these two neighborhoods for commute time to downtown"

That changes what matters in a listing. Specific, factual details, like lot size, garage, recent updates and commute distances, are what let an AI match a home to a request like "big backyard". Vague superlatives don't help. Zillow says results are sorted the same way as on its site and can be refined conversationally (Zillow).

Where the leads go

In both apps, the conversion happens on the portal, not in ChatGPT. Zillow says tour requests, agent connections and financing stay inside its own systems, and that OpenAI doesn't get MLS data for training (Real Estate News, sponsored content by Zillow). On Zillow, buyers who reach out are connected with Zillow-affiliated agents (HousingWire).

Realtor.com says its app is "designed to keep agents and MLSs at the center of the real estate transaction", that listing data isn't used for model training, and that it gets listings through direct MLS relationships rather than IDX (Online Marketplaces).

What this means for agents

1. Listing accuracy matters more. AI answers pull straight from listing data, so wrong square footage or stale status follows the listing everywhere.
2. Photos and descriptions are your ad. In a chat, buyers see a few images and a summary. Lead with your best photo, and make the first sentence count (within fair housing rules: see AI listing copy and fair housing).
3. Portal strategy is AI strategy. If you rely on portal leads, AI search sends more traffic through those same portals.
4. Listings must be on the MLS to show up. Zillow's standards exclude listings that are publicly marketed but not widely shared (our explainer).
5. Be the expert AI can't be. Chat tools answer "what can I afford?" They don't negotiate, read a seller, or spot a foundation problem.

Your own AI visibility

Buyers also ask ChatGPT and other assistants things like "best agent for first-time buyers in Austin." To be findable:

• Keep your Google Business Profile, website and portal profiles consistent
• Collect genuine reviews across platforms
• Publish specific local content, such as neighborhood guides with real data
• Make your service area and specialties explicit on your site

FAQ

Do buyers get connected to the listing agent?

On Zillow, buyers are routed into Zillow's own agent connection flow. Listing attribution shows the broker and MLS, but the inquiry path is Zillow's.

Is my MLS data training ChatGPT?

Zillow and Realtor.com both say listing data isn't used for model training.

Should I pay for portal leads because of this?

That's a business decision. AI search currently pushes traffic through the portals, so evaluate cost per closed deal, not cost per lead.

Changelog: Oct 13, 2026: first published.