According to Inc, Zillow built the first real estate app for ChatGPT in just six weeks after meeting with OpenAI executives before Sam Altman’s October DevDay announcement. The app uses OpenAI’s Apps Software Developer Kit and Anthropic’s Model Context Protocol to connect ChatGPT to Zillow’s property data. Users can search by typing commands like “Hey Zillow, find me 2 bed, 1 bath condos selling for $1 million in Brooklyn” or using the @ symbol to invoke the app. Zillow’s CTO David Beitel confirmed they maintained full control over their data and interface despite the tight timeline. The company worked through constant changes as OpenAI was still developing the framework, making adjustments right up until launch day.
Sponsored content — provided for informational and promotional purposes.
Why Zillow moved so fast
Here’s the thing: when you’ve got 800 million weekly users on the table, you don’t wait around. Beitel basically said Zillow had to be where their customers are, and right now that’s ChatGPT. But it’s not like they’re new to AI – they’ve been using machine learning since 2006 with their Zestimate system. So this wasn’t some panic move into unfamiliar territory.
What’s really interesting is how they managed the development chaos. Things were breaking constantly because OpenAI was still building the platform while Zillow was building their app. That’s like trying to build furniture while someone’s still designing the room around you. Yet they shipped in six weeks. That’s either incredibly impressive or slightly terrifying, depending on your perspective.
The bigger AI strategy
Zillow isn’t putting all their eggs in one basket. They’re using Google’s Gemini, OpenAI’s enterprise plan, and startup Glean internally. Beitel claims these tools have saved employees over 275,000 hours. That’s massive. But here’s where it gets really strategic – they’re not talking about replacing agents. They’re talking about creating “super agents.”
Think about that for a second. The real estate industry has been terrified of AI making human agents obsolete. But Zillow’s approach is different – enhance rather than replace. It’s smart positioning in a heavily regulated industry where human expertise still matters enormously.
What this means for real estate
The home buying process is ridiculously complicated. Beitel isn’t wrong – it involves multiple steps, tons of people, endless information, and takes months. That complexity makes it perfect for AI disruption. But here’s the catch: real estate is also heavily regulated and incredibly local. Can ChatGPT really understand neighborhood nuances? Zoning laws? School district boundaries?
Zillow’s playing the long game here. They got in first on ChatGPT’s new app platform, which gives them prime positioning. They have the brand recognition and the data to make this work. But the real test will be whether users actually stick around in ChatGPT or just use it as a starting point before switching to Zillow’s full website.
This feels like the beginning of something much bigger. If Zillow can make natural language property search work, it could fundamentally change how people start their home buying journey. No more fiddling with dropdown menus and sliders – just tell the AI what you want. The question is whether other real estate platforms will follow suit or get left behind.
