Comparable-sales valuations, returned in one request.
Send a suburb and a floor area, get back an estimated value, a price range and a confidence score — built from recent comparable sales, not list-price guesses. Quote before the manual CMA is even open.
Private beta — keys are issued on signup, no waitlist.
{
"suburb": "Claremont",
"city": "Cape Town",
"floor_area_sqm": 180,
"estimated_value_rands": 2850000,
"value_range": {
"low_rands": 2690000,
"high_rands": 3010000
},
"price_per_sqm_rands": 15833,
"confidence": { "score": 0.81, "label": "high" },
"comparables_used": 14,
"method": "comparable_sales"
}A real call, not a mockup
The example above isn't illustrative — this is the same GET /v1/valuation call, captured live against a running instance of the API through its interactive docs.
Claremont · 180m² · 3-bed house · 12 comparables 200 OK — confidence 0.98 (high)
From raw sales to a defensible number
Every valuation runs the same four-step pipeline — no black box, and the response tells you exactly how many comparables and how much confidence went into it.
Filter comparables
Sales within the lookback window, matching property type, within one bedroom, and inside a sane price-per-sqm band for the area.
Remove outliers
An interquartile-range pass drops the sales that would otherwise skew the estimate — a distressed sale or an outlier renovation doesn't move the number.
Weight by recency
The median price per square metre is skewed toward more recent sales, so a suburb that's moved in the last quarter shows it.
Score confidence
A 0–1 score blended from comparable count, price variance and recency — so a thin dataset reads as thin, not confident.
Everything a valuation needs, nothing it doesn't
Suburb, city and property specs in — an estimated value, price range, price/sqm and confidence score out.
The raw comparable sales behind any estimate, for when an agent wants to show the evidence, not just the number.
Median price, price per square metre and growth for a suburb, refreshed nightly rather than computed per request.
Real sales, not list-price guesses
Estimates are built with the comparable sales method — the same approach an agent uses for a CMA, run automatically against a database of South African property sales and listings, kept current on an ongoing basis rather than a one-off import.
Every response reports its own comparable count and confidence score. An estimate built on two thin comparables says so, rather than presenting the same confidence as one built on twenty — so you know when to hand it to a client as-is, and when to double-check it first.
Priced by call volume
Every plan includes all three endpoints. Upgrade or downgrade any time — a new key is emailed within seconds of checkout.
Billing via Lemon Squeezy. Prices in ZAR, ex. VAT where applicable.