Why Adresse.ai exists

Adresse.ai was inspired by one family’s frustration and confusion with the French property-buying process. They had spent fifteen years using Zillow, Redfin, and Rightmove. They had taken instant access to comparable evidence for granted. They sat down on a Tuesday evening to look up what similar properties had sold for near the village they were considering. Three hours later, they had downloaded the Demande de Valeurs Foncières CSV from data.gouv.fr, written enough Python to filter for vaguely relevant properties, and hand-looked-up the DPE on each comparable from the ADEME register.

That was the night the missing tool went from concept to fact.

The process is harder than it should be

Buying property in France, as an buyer, is a particular kind of hard. The legal infrastructure is excellent. The notaires are competent. The data is, on paper, more open than the UK or US equivalents. None of that helps when you’re standing in front of a stone mas near Uzès, an asking price of €685,000 in your hand, and a French agent who is professionally honest but paid by the seller. You don’t know whether the price is fair. You don’t know whether the comparables on the same village square sold for €600 per square metre or €4,000. You don’t know what the energy rating is going to cost to fix. You don’t know whether the asking price has held flat for fourteen months because the seller is patient or because the price is wrong.

In the US or UK you’d type the address into a portal and have an answer in fifteen seconds. In France, in 2025, you trusted the asking price, you trusted the agent, or you spent a Saturday afternoon in a French-language CSV file. None of those three paths is good.

What’s actually missing

France has been quietly running one of the most open property-data systems in Europe since 2019. DVF is comprehensive, free, and authoritative. ADEME publishes every energy rating. Notaires de France produces serious quarterly market analysis. The cadastre is queryable. The legal infrastructure that took the UK and US a generation of regulation to build, France has had since the Loi ESSOC shipped.

What France doesn’t have is a tool that translates this data into an answer for a buyer who doesn’t read French and doesn’t have the patience for CSV files. MeilleursAgents is built for sellers in French. Patrim requires a French tax account most pre-purchase buyers don’t have. PriceHubble is sold to banks. The English-language editorial sites publish guides, not numbers.

So a buyer researching French property in 2025 paid the asking price, or trusted the agent, or asked Reddit. Those three options were what stood between millions of buyers and a half-million-euro decision.

That’s a bad equilibrium.

What changes

Three things, modestly:

Buyers walking into French listings should arrive armed with the same data transparency they take for granted at home. Right now, they arrive disarmed. That can be fixed.

The asking price in France should stop being treated as the market price. It isn’t, the data is unambiguous on the gap, and the buyers who don’t know to check the gap are paying for that ignorance. The gap should narrow as more buyers carry comp pools into negotiations.

The agent-buyer relationship should evolve toward what it is in the UK and US: professional, useful, and not the only source of market data. The agent’s job is to find and represent. The data tool’s job is to be the data tool. France’s property market is healthy enough that more transparency makes the whole system work better.

What Adresse.ai isn’t

Not a Zestimate replacement. The Zestimate has known flaws in its own market; copying its model in a different country would be worse. Adresse.ai is closer to “the kind of comparable-sales transparency US and UK buyers know, in a market that didn’t have it, with the actual government data made readable.” That framing is honest. The full reasoning is at /guides/zestimate-for-france.

Not a brokerage, a chasseur immobilier, or a French lawyer. Adresse.ai helps you decide. Other people execute. The discipline of staying narrow is the only way the product stays useful.

Not finished. The MVP launches with the South of France because that’s where the data and tuning hold up. Other regions land as the analysis lands, one at a time, paired with real product capability rather than ahead of it.

Try it on your listing

If you have a French listing in front of you, paste it. The frustration that produced Adresse.ai is exactly the frustration the tool removes.

Run an estimate →

See also: