As of Q2 2026, San Francisco's housing market is moving quickly: the citywide median sale price is $1.83M at a median 12 days on market, with closings averaging $1,225 per square foot. The AI and tech hiring cycle is the demand story behind that speed, concentrating buyers in the city rather than dispersing them.
Unless otherwise noted, all market figures in this report are drawn from MLS data, measured over the trailing twelve months as of Q2 2026. Start with the number that frames everything else: the citywide median close sits at $1.83M, and homes are trading in a median of 12 days. That is a fast market by any historical standard for San Francisco. The average sale price is higher, $2.41M, which tells you the top of the market is doing heavy lifting on the average while the median holds the middle. The point of this report is to connect those numbers to the force most people are asking about right now: the AI boom, and whether it is actually changing what homes cost and how fast they sell in this city.
San Francisco by the numbers: Q2 2026
What the speed of the market is telling us
The single most useful signal in this data is not the price, it is the 12-day median days on market. As of Q2 2026, half of San Francisco's closed sales went under contract in less than two weeks. A 12-day median is the signature of demand outrunning supply: buyers who are ready, financed, and willing to compete, against an inventory base that has not expanded to meet them. When a market sits this tight, the headline price is a lagging indicator. The leading indicator is how long a listing stays available, and right now it does not stay available long.
I read the $1.83M median and the $2.41M average together, because the gap between them carries information. When the average runs well above the median, the upper tier of the market is transacting actively enough to pull the mean up. That pattern is consistent with a city where high-compensation buyers are present and bidding, which is exactly what you would expect if a hiring cycle were feeding demand at the top.
How the AI boom actually shows up in housing
Here is the honest version, because the AI-and-real-estate story attracts a lot of confident numbers that do not survive scrutiny. I am not going to quote you a precise figure for AI jobs created or dollars of compensation flowing into the city, because that is not data I can source cleanly, and a number nobody can stand behind is worse than no number at all. The useful question is how a hiring cycle like this transmits into a housing market, and what to watch for as it plays out.
A concentrated employment boom in a supply-constrained city does three things to housing. It pulls in-migration toward the metro, because the jobs are physically here and a meaningful share of AI work has gravitated back toward in-person collaboration. It raises the share of buyers who can compete on cash or on speed, because the compensation in this sector skews high and often includes equity. And it does almost nothing to the supply side in the near term, because San Francisco cannot add housing on the timeline that hiring moves. Demand can re-rate in a quarter. The building stock cannot. That asymmetry is the whole mechanism, and it is why a tight days-on-market reading is the first place the boom becomes visible.
Where it does not show up cleanly is in any single neighborhood you can point to and call the AI neighborhood. The demand is diffuse. It lands in the close-in condo corridors near the job centers, in the family-scale houses on the west side for buyers a few years into their earnings, and at the very top for the founders and early employees who have liquidity events. Treating the boom as a localized phenomenon misreads it. It is a citywide pressure on a citywide shortage.
The other thing worth naming is the timing mismatch between an earnings event and a home purchase. A hiring boom does not convert to closed sales the moment the offer letters go out. Equity vests over years, bonuses arrive on schedule, and most buyers wait until they are confident the role and the compensation are durable before they commit to a multi-million-dollar mortgage. That lag means the demand pressure a hiring cycle creates tends to build gradually and then persist, rather than spiking and fading. It also means the speed you see in the Q2 2026 data reflects hiring decisions made over the preceding period, not just this quarter, which is part of why a tight days-on-market reading can hold even when the headlines about any single company cool off.
What to consider before you read the boom into your offer
The macro story is real, but it is dangerous to translate it directly into a bid. Here is where the AI-boom narrative misleads if you take it at face value.
A hiring cycle is not a price floor
The boom supports demand, but it does not guarantee that any specific home holds or grows its value. Employment cycles turn. The buyers I work with who do best are the ones who underwrite the property on its own merits, comps, condition, location, and treat the macro tailwind as context rather than as a reason to overpay. If the only thing supporting your number is a headline about AI, your number is fragile.
The citywide median hides two different markets
A $1.83M median blends a sub-$1M condo with a multi-million-dollar house. As of Q2 2026 those segments are not moving on the same clock, and the boom touches them differently. Price within one property type and one part of the city, never against the citywide figure as a whole, which is useful for a headline and useless for an offer.
Speed cuts both ways
A 12-day median days on market is a seller's signal, but it is also a discipline test for buyers. In a market this fast, the impulse is to skip inspections and contingencies to win. I do not sugarcoat this part: that is how people overpay and inherit problems. The win is to be fully prepared so you can move in 12 days without abandoning the protections that keep a fast purchase from becoming a regret.
What's driving the numbers
Two forces explain the Q2 2026 readings. The first is structural scarcity. San Francisco's inventory has not grown to match the demand sitting in front of it, and a 12-day median days on market is the direct evidence. When homes clear that fast, supply is the binding constraint, and price firmness follows almost automatically. The second is the employment backdrop. A high-compensation hiring cycle, of which the current AI wave is the most visible part, raises both the number of qualified buyers and their willingness to compete. I would not overstate the precision here, because employment effects on housing are real but slow to measure cleanly. The pattern in the data, fast sales and an average well above the median, is consistent with that demand pressure rather than with a softening market.
What to watch through the rest of 2026
San Francisco runs two real selling cycles, spring through the end of June and a shorter fall window from late August to Halloween. Into that fall window, watch three things. First, whether inventory loosens at all; with a 12-day median days on market as of Q2 2026, supply is the story, and any meaningful increase in listings would cool the speed before it cools the price. Second, whether the hiring cycle holds; the AI demand tailwind is a tailwind only while the jobs keep landing here, and that is genuinely unknowable quarter to quarter. Third, the broader rate environment, which sets how much of that demand can convert to closed sales. If current patterns hold, the data suggests continued firmness into the fall, but a forecast is not a guarantee, and anyone who tells you the boom can only push prices one direction is selling certainty that does not exist.
Is the AI boom driving up San Francisco home prices in 2026?
It is supporting demand rather than directly setting prices, as of Q2 2026, when the citywide median sits at $1.83M and homes trade in a median of 12 days. A high-compensation hiring cycle raises the number of competitive buyers in a city that cannot quickly add supply, and the clearest fingerprint of that pressure is speed, the 12-day median days on market, more than any single price figure you can attribute to AI alone.
What is the median home price in San Francisco right now?
The citywide median sale price is $1.83M as of Q2 2026, measured over the trailing twelve months, with an average sale price of $2.41M and a median of $1,225 per square foot. The gap between the $1.83M median and the higher average reflects an active luxury tier pulling the mean upward. Treat the median as a starting frame, not as a valuation for any specific property type or part of the city.
Are AI jobs bringing buyers back to San Francisco?
The market behavior is consistent with returning, competitive demand as of Q2 2026, when the median days on market is just 12. A concentrated hiring cycle pulls workers toward the metro because the jobs are physically here, and that shows up first as faster sales and tighter competition. Attributing a precise share of buyers to AI specifically would overstate what the data can prove, because that figure is not cleanly sourceable, but the directional pull is visible in the speed.
Is it a good time to buy in San Francisco during the AI boom?
It can be, as of Q2 2026, but only if you buy with discipline in a 12-day-median market. The boom supports demand, which means competition is real and preparation wins; being fully financed and decisive matters more than timing the cycle. The buyers I work with do best when they underwrite the home on its own comps and condition and treat the AI tailwind as context, not as a reason to skip protections or overpay.
Will the AI boom cause a San Francisco housing bubble?
That is not something the current data can confirm or rule out as of Q2 2026, when the citywide median is $1.83M and sales are fast at 12 days. A demand-driven, supply-constrained market can stay firm for a long time, but employment cycles turn, and a market leaning heavily on one hiring wave carries concentration risk. The honest answer is to watch inventory and the hiring pace rather than assume the tailwind only blows one way.
Talk to me about your block and price tier
The numbers in this report are the citywide view, and the AI-boom story is a macro frame. Your block, your property type, and your price tier behave differently, and that is the level where an offer actually gets won or lost. If you want to see what is moving on the ground, you can browse current San Francisco listings or read the latest Central Richmond Real Estate Market Report for a neighborhood-level read. When you are ready, I can pull the comp set for your specific situation and walk you through what this market means for buying or selling this season.
Last updated: June 30, 2026
