French Property Data API: The Complete Guide for Developers

If you're building a real estate application for the French market, you'll quickly discover that getting property data is harder than it looks. Unlike the United States, where the MLS provides a centralized feed of listings, France has no equivalent system. Property data is fragmented across hundreds of portals, agency websites, and classified platforms — each with its own format, its own identifiers, and its own quirks.

This guide explains the French real estate data landscape, the technical challenges of working with it, and how to access structured, deduplicated French property data through an API.

The French Real Estate Data Landscape

France's property market operates without a centralized listing database. Every portal, every agency network, and every classified site runs independently. There is no shared property identifier, no common data standard, and no universal feed that aggregators can subscribe to.

Here are the major players:

LeBonCoin — France's largest classified advertising platform. It dominates in volume because it accepts listings from both agencies and private sellers. That breadth is also its weakness for data consumers: listings from private individuals tend to have less structured data, inconsistent descriptions, and lower geocoding accuracy than agency listings.

SeLoger — The largest dedicated real estate portal in France. SeLoger has a strong agency presence and generally higher data quality per listing than LeBonCoin. It's the first place most French buyers go when searching through an agent.

BienIci — An aggregator backed by a consortium of real estate software providers (including major French CRM vendors like Apimo, Hektor, and AC3). BienIci receives listings directly from agency management software, which means its data tends to be well-structured. However, its coverage depends on which software providers participate in the consortium.

Logic-Immo — A long-established portal, now part of the SeLoger group (owned by Axel Springer). It still operates as a separate platform with its own listing base, but there's increasing overlap with SeLoger's inventory.

PAP (Particulier à Particulier) — Focuses exclusively on direct sales and rentals without an agent. PAP listings represent a distinct segment of the market — typically priced lower (no agency fees) but with less professional presentation and data quality.

Beyond these five, there are hundreds of individual agency websites, regional portals, franchise networks (Century 21, Orpi, Laforêt, Guy Hoquet), and notaire listing platforms. Notaires — the notaries who handle all French property transactions by law — maintain their own databases of properties for sale, particularly in rural areas where agency presence is thin.

Government Open Data: DVF

France does have one valuable open data source: DVF (Demandes de Valeurs Foncières), published on data.gouv.fr. DVF records every completed property transaction in France, including the sale price, property type, surface area, and location.

DVF is genuinely useful for historical price analysis and valuation benchmarking. But it has clear limitations:

Delayed by months. Transactions appear in DVF only after they're registered by the tax authority. Expect a 3-6 month lag between a sale closing and the data appearing in DVF.
No active listings. DVF only covers completed transactions. It tells you what properties sold for, not what's currently on the market.
No photos, descriptions, or contact information. It's a transaction record, not a listing.
No rental data. DVF covers sales only.

For developers building applications that need current market activity — active listings, rental prices, new-to-market properties — DVF alone is insufficient. It works best as a complement to live listing data, not a replacement for it.

Why French Property Data Is Hard

If you've worked with US MLS data or UK property feeds, the French market will feel like a different world. Here are the specific technical challenges:

No Standard Identifiers

The same apartment has a different ID on every portal. LeBonCoin assigns one, SeLoger assigns another, and the agency's own CRM has a third. There is no French equivalent of a parcel number or MLS ID that follows a property across platforms. Matching listings across sources requires fuzzy matching on address, surface area, price, and number of rooms — which is exactly as error-prone as it sounds.

Duplication Everywhere

A typical agency workflow: the agent enters a listing in their CRM software, which pushes it to SeLoger, LeBonCoin, BienIci, Logic-Immo, and the agency's own website. That's five copies of the same property, each with slightly different formatting, sometimes different photos, and occasionally different prices (some portals include agency fees in the displayed price, others don't).

If you're aggregating data from multiple sources without deduplication, your dataset will dramatically overcount the actual number of properties on the market.

Inconsistent Data Quality

Geocoding quality varies wildly across sources. Some listings include exact GPS coordinates accurate to the building level. Others provide only a city name or postal code, with coordinates pointing to the city center. Surface area might appear in a dedicated field on one portal and buried in the description text on another. Room counts follow different conventions — some portals count kitchens as rooms, others don't.

Energy performance certificates (DPE — Diagnostic de Performance Energétique), which are mandatory for all French property listings, are formatted differently across portals. Some provide the actual energy consumption value in kWh/m²/year, others just the letter grade (A through G), and some provide both but in different fields.

Scraping Is Fragile and Legally Risky

French portals update their front-end code regularly. A scraper targeting LeBonCoin or SeLoger that works in January will likely break by March. Both platforms use anti-bot measures including CAPTCHAs, browser fingerprinting, and rate limiting.

More importantly, scraping these portals typically violates their terms of service. French courts have ruled on web scraping cases, and the legal landscape under GDPR adds another layer of complexity when the scraped data includes personal information (agent names, phone numbers, email addresses).

No Real-Time Push Feeds

US MLS systems push updates to authorized consumers via RETS or Web API feeds. French portals don't offer anything comparable. To detect new listings or price changes, you need to actively monitor each source — polling at regular intervals and diffing the results against your previous snapshot.

DIY vs. API: The Build-or-Buy Decision

Every team building a French property data product faces this question: do we build our own data pipeline, or do we use an existing API?

Building Your Own Pipeline

The arguments for building in-house:

Full control over which sources you monitor and how you process the data
No ongoing API subscription costs
Ability to customize deduplication logic for your specific use case

The reality:

Expect 6+ months of engineering time before you have a reliable, multi-source pipeline producing clean data. That's 6 months before your product can even start using the data.
Plan for 1-2 full-time engineers dedicated to maintaining scrapers, fixing breakage, updating parsers, and monitoring data quality. This is not a "build it once" system.
Legal review of your scraping practices will add cost and potentially limit which sources you can target.
Deduplication across sources is a hard problem. Getting it wrong means either missing properties (false positive matches) or double-counting them (false negative matches). Both undermine your product's credibility.

Using a Data API

The arguments for an API:

Structured, deduplicated data from day one. Your team can start building product features immediately instead of spending months on data infrastructure.
Source monitoring, parsing, and deduplication are someone else's problem.
The API provider handles legal compliance with data sources.

The trade-offs:

Ongoing subscription cost.
Dependency on a third party for a core part of your stack.
Less control over exactly how data is collected and processed.

For most teams, the API route makes sense unless data infrastructure is your core product. If your competitive advantage is a better search experience, a smarter valuation model, or a faster alert system — spending your engineering budget on data plumbing is a poor allocation.

Working with French Property Data via API

Stream.estate aggregates property data from 900+ French sources — major portals like LeBonCoin, SeLoger, BienIci, Logic-Immo, and PAP, plus hundreds of individual agency websites, franchise networks, and notaire listings. New listings appear in the API within an hour of publication on the source portal. The platform ingests over 50,000 new listings daily.

Deduplication Across Sources

The most immediate value of the API is deduplication. When the same three-bedroom apartment in Lyon's 6th arrondissement appears on LeBonCoin, SeLoger, the agency's own website, and BienIci, Stream.estate identifies it as a single physical property. You get one result per property, with each source listed as a separate advert under that property.

This alone saves significant development time. Without it, you'd either need to build your own deduplication engine or accept inflated listing counts that mislead your users.

Searching Properties

A straightforward search for apartments for sale in Lyon, with results deduplicated across all sources:

# Search apartments for sale in Lyon
curl "https://api.stream.estate/documents/properties?transactionType=sale&propertyType=apartment&city=Lyon&itemsPerPage=10" \
  -H "X-API-KEY: your-api-key"

You can filter by property type (apartment, house, land, parking, commercial), transaction type (sale, rent), price range, surface area, number of rooms, and geographic criteria (city, postal code, department, or bounding box coordinates).

Authentication uses a simple API key passed in the X-API-KEY header. No OAuth flows, no token refresh logic.

Market Indicators

For market analysis, the price-per-meter endpoint provides aggregate pricing data by location:

# Get price per square meter for sales in Lyon
curl "https://api.stream.estate/indicators/price-per-meter?city=Lyon&transactionType=sale" \
  -H "X-API-KEY: your-api-key"

This is particularly useful for building automated valuation models, market comparison tools, or investment analysis dashboards. You can break down pricing by city, department, or postal code to identify local price variations — which in French cities can be dramatic. In Paris, the price per square meter can differ by a factor of two between neighboring arrondissements.

Similar Properties

Finding comparable properties is essential for valuations and market positioning. The API provides a similar-properties endpoint that returns listings comparable to a given property based on location, type, surface area, and price.

Points of Interest and City Data

Property value in France is heavily influenced by proximity to public transport (especially Métro stations in Paris and Lyon), schools, and local amenities. The API includes POI (points of interest) data and city-level statistics that help contextualize a property's location.

Common Use Cases in France

Based on how developers actually use French property data, these are the most common applications:

Property search portals. Build a search engine covering the entire French market without maintaining 900+ scrapers. Whether you're building a general-purpose portal or a niche product (luxury properties, student rentals, commercial spaces), the underlying data call is the same — you just filter differently.

Automated valuations (AVM). Combine current listing prices from the API with DVF's historical transaction data to build valuation models. The API gives you what's on the market right now and at what asking price. DVF tells you what similar properties actually sold for. Together, they provide a solid foundation for price estimation. For a step-by-step implementation, see our tutorial on building a property valuation tool with an API.

Market analysis and reporting. Track price trends by city, department, or postal code over time. Identify markets where listing volume is increasing or decreasing. Compare rental yields across different French cities by combining sale prices with rental listing data.

Lead generation for agents. Monitor new listings matching specific buyer criteria and notify agents in real time. The API supports webhooks, so you don't need to poll — you get notified when a matching listing appears.

Investment analysis. Compare rental yields across different French cities and property types. Identify undervalued areas where the gap between listing prices and comparable transaction prices suggests room for negotiation.

Price alerts. Notify users when properties matching their search criteria drop in price or when new listings appear in their target area. With sub-hour data freshness, these alerts arrive before most manual search processes would catch the same changes.

Getting Started

Get an API key at stream.estate. The signup process gives you immediate access to the API.
Start with a count query. Set itemsPerPage=0 to see how many results match your filters without retrieving the actual listings. This helps you understand data volume before building pagination logic.

# Count apartments for rent in Paris
curl "https://api.stream.estate/documents/properties?transactionType=rent&propertyType=apartment&city=Paris&itemsPerPage=0" \
  -H "X-API-KEY: your-api-key"

Build your filters. Narrow by city, property type, transaction type, price range, surface area, or any combination. Test different filter combinations to understand coverage in your target market.
Set up webhooks for real-time notifications on new listings or price changes matching your criteria. This is more efficient than polling the API on a schedule.
Combine with DVF for historical context. Use API data for current market activity and DVF for completed transaction prices. The two datasets complement each other well.