Our Data & AVM
The Gold Standard in Real Estate Analytics + Valuations
Validated every quarter by independent third-party testing. Powered by a machine learning-based methodology and direct, real-time data from more than 300 MLSs and over 3,000 county assessor offices, from a licensed real estate brokerage operating nationwide.
Valuations built on up to 35 years of recorded sales history
Built on census block-level detail
National coverage
Figures from HouseCanary's internal purchase benchmark, six months ending Aug. 18, 2026.
How it works
Built on valuation science
Comprehensive data plus machine learning. Every valuation follows the same three steps, run for every census tract in the country.
Gather and clean the data
Property records from 3,100+ county assessor offices, recorded sales from 2,700+ county recorder offices and real-time MLS data, refreshed daily. Every source is checked field by field, and anything suspicious is quarantined until it's verified.
Build block-level price indices
Machine learning builds home price indices for 3.4 million US census blocks, by property type. Every price observation is adjusted to today's value.
Tune the valuation models
Multiple machine learning models estimate each home's value from property, neighborhood, economic and repeat-sale data. Their estimates are combined into one value, tuned to minimize bias and maximize accuracy.
Pre-list valuations methodology
Pre-list optimized
HouseCanary produces 120m+ valuations each month, calibrated for pre-list accuracy. It was among the first firms to implement this approach and consistently ranks among the industry's best by third-party evaluations.
Block-level detail
The census block is the smallest group of properties in the U.S. Modeling at that level captures how prices move street by street, even in blocks with few sales.
Machine learning-based
Machine learning finds complex relationships classic models miss, using every known past sale rather than just recent comps. Models are recalibrated monthly and update instantly when a new listing or sale arrives from a direct MLS feed.
Every value comes with a confidence score
Each valuation includes a forecast standard deviation (FSD), a range built to capture the eventual sale price about 68% of the time. An FSD below 0.15 marks a high-confidence value.
Client proof
Trusted by the biggest, best
Large financial institutions such as mortgage lenders, secondary market issuers and investors rely on HouseCanary's industry leading property intelligence and valuations to price and fund mortgages, value portfolios, manage risk and make better, faster decisions. They choose HouseCanary for estimates that come with a clear confidence score, data protected against manipulation, regular independent testing and the documentation their risk and compliance teams need.
What it enables
Better data. Better decisions.
Accurate valuations
HouseCanary values off-market homes within a median 6.8% of their eventual sale price, ranking it among the top-performing AVMs based on third-party testing.
Faster pricing
MLS data updates as often as every 15 minutes, and valuations refresh the moment a new listing or sale comes through.
Unique opportunities
Market intelligence few providers can match, like the weekly Market Pulse, which tracks up to 27 key metrics for every single-family home, condo, and townhouse on the market.
The gold standard
What sets HouseCanary apart
Depth + breadth
HouseCanary is a national real estate brokerage that is also a data and valuation provider. Its participation in a majority of US MLSs gives it unmatched depth and breadth of real-time real estate data, valuations and transactions.
Data security + compliance
HouseCanary maintains the highest data standards, with annual SOC 2 Type II audits. It's a platform trusted by the largest, most rigorous businesses in the financial, technology and real estate world.
Pre-list optimized
HouseCanary's AVM is built with a pricing methodology before homes are listed, using full county property sales history going back 20 years, at the census block level. The pre-list valuation is particularly valuable when trying to accurately price and value off-market homes.
Compliance
Built to federal AVM quality control standards
Federal regulators, including the FHFA, require AVMs used in mortgage lending to meet five quality control standards, effective Oct. 1, 2025. Here's how HouseCanary meets each one.
| Standard | How HouseCanary meets it |
|---|---|
| 1High confidence | Property-level confidence scores, with high confidence defined as FSD below 0.15. |
| 2Data integrity | Field-by-field validation, anomaly quarantining and monthly model recalibration. |
| 3No conflicts of interest | Values are never modified by, or contingent on, any transaction or credit decision. |
| 4Random sample testing | Internal testing weekly, plus quarterly testing by two independent third-party firms. |
| 5Nondiscrimination | Monthly disparate impact testing, and no protected attributes or proxies like school scores in the model. |
Solutions
Put the gold standard to work
For Real Estate Investors
Underwrite acquisitions, value portfolios and spot opportunities with property-level data and valuations.
Learn MoreFor Real Estate Professionals
Price listings, win offers and grow business with CMAs, market intelligence and seller prospecting.
Learn MoreFor Mortgage Lenders
Make faster, more confident lending decisions with AVMs built to align with federal quality control standards.
Learn MoreCitations
- HouseCanary Valuation Whitepaper, August 2026. Data sources, methodology, forecast standard deviation and testing. National median absolute prediction error of 2.50% on 1,898,015 transactions, six months ending Aug. 18, 2026. Request it in our Trust Center
- HouseCanary's Compliance with FHFA Standards for AVM Use in Financial Institutions, memo, Aug. 18, 2026. Alignment with federal AVM quality control standards, data sources, SOC 2 Type II audits, national benchmarking and disparate impact testing. Request it in our Trust Center
- Quality Control Standards for Automated Valuation Models, final rule, OCC, Federal Reserve Board, FDIC, NCUA, CFPB and FHFA, effective Oct. 1, 2025. Read the final rule
AVM Disclosure: An AVM is an estimated sale price for a property. It is not the same as the opinion of value in an appraisal developed by a licensed appraiser under the Uniform Standards of Professional Appraisal Practice.