San Francisco residential real estate, district by district. Single Family + Condo / Townhouse. Presented Jun 1, 2026.
Volume is roughly normal (484 closes vs the 488 ten-year May average) but listings are at a trough. The result: prices at decade highs, 83% of Single Family homes sold above list, and cash competition climbing. At $5M+, 64% of Single Family sales close in cash, and Condos run cash-heavier than Single Family at every band above $1M.
Citywide sales +10% YoY (484 closes vs 439 in May 2025), and +7% vs April. Volume is normal. The inventory gap (691 vs 1,163 active) is what is keeping competition tight.
Each cell shows median Sale vs List (big number), number of closes, share that sold above list, and median price per square foot. Shade tracks heat intensity. Original list-price band, not final.
| LIST PRICE → | Under $1M | $1–$1.5M | $1.5–$2M | $2–$3M | $3–$5M | $5M+ |
|---|---|---|---|---|---|---|
| SINGLE FAMILY 247 closes |
100% $771 / sqft |
122% $958 / sqft |
134% $1,137 / sqft |
127% $1,185 / sqft |
122% $1,422 / sqft |
106% $1,997 / sqft |
| CONDO + TOWNHOUSE 237 closes |
100% $903 / sqft |
102% $1,120 / sqft |
103% $1,273 / sqft |
113% $1,424 / sqft |
121% $1,778 / sqft |
137%* $2,062 / sqft |
* Condo $5M+ band had only 2 sold listings, directional only. Original list-price (SF asks are often strategic underprices, so the band reflects what sellers signaled, not what buyers paid). Source: SF MLS via InfoSparks. Tenancy in Common excluded. Sales above 200% Sale-vs-List excluded as data-entry outliers.
Price trend smoothed to remove seasonal noise. Cash share broken down by property type, year-to-date.
Charts come from the AI-backed workflow Paulo built. It reads the latest MLS data, applies the cuts and filters Joske uses, and draws the same view every cycle.
Each card shows the current sold count, the Single Family vs Condo / Townhouse split with YoY change, the district’s overall Sale-vs-List, and the top sale.
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
Property details by Subdistrict
YoY = May 2026 vs May 2025. * = thin sample (fewer than 5 closes); treat the number as directional only. Single Family district medians can swing on small n, especially in lower-volume districts. Source: SF MLS via InfoSparks. TIC and outliers above 200% Sale-vs-List excluded.
This report is the citywide picture. Your block, your floor, your timing. Those tell a different story. A 15-minute conversation usually surfaces what a static report cannot.
Joske decides what gets tracked and how it gets read. The price bands, the comparisons, the filters, the outlier rules, what counts as a trend versus noise: those are his calls, drawn from twenty years of San Francisco transactions.
He looks at every cycle before it ships. If a number does not pass his read, it does not go out.
Paulo built the AI-backed workflow that turns Joske’s rules into the same report every two weeks. The workflow pulls the latest MLS data, applies the filters Joske uses, and builds the charts, the deck, and this website. It handles the spreadsheet work so Joske can focus on the read.
The point is not the AI. The point is that Joske’s rules are written down once and run the same way every cycle, which keeps Joske in his zone.
For another angle on the SF market, built with the same kind of AI-backed workflow: real-estate-in-the-ai-era.netlify.app
Sources: SF MLS via InfoSparks for Active, Pending, and Sold counts. Compass for off-market and private exclusives. Tenancy in Common excluded. Sale-vs-List above 200% excluded as data-entry outliers. Cash percentages use closes with reported financing only. Cadence: bi-weekly, with monthly and quarterly editions when the cycle calls for a deeper read.