From First Number
to Final Score.

You start by telling the AI your income, existing debts, target price, and down payment. HomeRates.AI pulls the live 30-year conforming rate directly from the Federal Reserve (FRED) and runs a deterministic PITI calculation — principal, interest, taxes, and insurance — against your gross income and existing debts. L1 is scored using a deterministic formula by loan type: Conventional (LTV ≤80 → 85, ≤85 → 78, ≤90 → 70, >90 → 60), FHA (LTV ≤90 → 72, ≤95 → 65, >95 → 58), VA (LTV ≤80 → 88, else → 78), Jumbo (LTV ≤75 → 86, ≤80 → 80, else → 72). This reflects real-world lending risk tolerance, not an opinion.

Once a property address is in context (from any of the 3 entry points), Grok 4 scores the specific property against your qualified monthly budget. Three factors drive the score: (1) PITI vs. your qualified payment — the gap between what you're approved for and what this property actually costs month-to-month; (2) List price vs. Redfin/Zillow AVM — how the asking price compares to the automated valuations built from recent comp sales within 0.5 miles; (3) Days on market positioning — a property sitting 60+ days with a price cut is priced differently than a fresh listing. A property that fits the budget and prices at or below AVM scores 80+.

Grok 4 runs a live web search across the local market to score broader conditions independent of the individual property. Three signals determine the score: (1) Median days on market — a DOM above 60 days indicates buyer leverage; under 15 days means active competition; (2) Sale-to-list ratio — offers closing below 100% of list price signal negotiating room; above 102% means waived contingencies; (3) Recent comp velocity — how quickly similar properties are moving. Seller's markets with sub-20 DOM and 101%+ sale-to-list score 30–50. Buyer's markets with 60+ DOM and sub-98% sale-to-list score 70+.

Tapping Run Deep Analysis inside the Property Intelligence page triggers Grok-4 deep analysis via the xAI Responses API. Grok 4 searches the live web and synthesises seven sub-dimensions: Walk Score, Transit Score, Bike Score, Schools (GreatSchools ratings), Safety, Amenities & Commute, and Wildfire Risk. Wildfire Risk uses an inverted scale — a high wildfire danger produces a low sub-score, which pulls the composite down significantly. This is the signal most buyers skip and the one that most determines resale value, insurance premiums, and quality of life over time.