ChatGPT will happily tell you what a suburb's median house price is. It just won't tell you where it got it, when it was true, or whether it made it up. As of this week, it doesn't have to guess.
Microburbs is now an app in ChatGPT. Add it once, and when you ask about an Australian suburb or property, ChatGPT pulls the answer from our live data instead of from whatever it half-remembers from its training.
How do I add Microburbs to ChatGPT?
Thirty seconds, no install:
- Open ChatGPT and go to Plugins in the sidebar, or tap + in any chat.
- Search Microburbs .
- Click + to add it. That's it.
Or go straight to the listing: Microburbs in ChatGPT.
What can I ask it?
Anything you'd ask a buyer's agent who actually had the numbers in front of them. We ran these through it on 21 September:
- "What are median house and unit prices in Subiaco, WA?" — houses $2,220,000, units $830,000, with the date the figure was current.
- "How does crime in St Kilda compare with the national average?" — 227% above the national median, broken down by crime type.
- "What rental yields can investors get in Fortitude Valley?" — 5.01% gross for units, with vacancy, days on market and five-year growth alongside.
- "Which school catchments cover Unley, and how do the schools rank?" — the two public primary catchments, their NAPLAN rankings, and the private schools nearby.
- "What bushfire and flood risk is there in Sandy Bay?" — how much of the suburb sits in a bushfire-prone area, and how much under a flood overlay.
Every number in those answers came from a query against our database, made after the question was asked. Not a memory. Not a vibe.
Why not just ask ChatGPT on its own?
Because a language model doesn't know your suburb. It knows what was written about your suburb on the internet before it was trained — which might be a two-year-old news article, a forum argument, or a median for the wrong property type. And it will state any of those with exactly the same confidence.
That's a problem when the decision is a seven-figure one. Microburbs has been building Australian property data for over a decade, with over 5 billion data points covering every address, street and suburb in the country. Where most property data stops at the suburb, ours goes down to the street and the ABS mesh block — pockets of roughly 30 to 60 homes — because a suburb average hides most of what actually matters when you're choosing a house.
Plugging that into ChatGPT means the conversation you were already going to have now has real numbers under it.
Is it free?
Yes, to start. Without an account, the app answers for nine sample suburbs covering every state and territory: Bondi, Belmont North, St Kilda (Vic.), Fortitude Valley, Subiaco, Unley, Sandy Bay, Kambah and Nightcliff. Ask about anywhere else and it tells you straight, instead of inventing a figure.
For every other suburb and address in Australia, there's the full Microburbs API — the same data, priced per call, which plugs into Claude Code, Cursor, your own agents or your own product. Sign-in from inside ChatGPT is on our list.
Now go and ask it about your street. It's a better conversation than the one you were having.