Airbnb grew Q2 2026 revenue 17% to $3.61B and lifted operating margin to 21.0%, but 27% net income growth shrinks to about 15% once a one-off $77M tax benefit and a 4.6% smaller share count are stripped out.
Revenue up 17%, but a third of the earnings beat came from tax and buybacks
Airbnb's second quarter of 2026 (the three months ended June 30, 2026) was a genuine operating improvement wrapped in a flattering headline. Revenue rose 17% to $3.608 billion, and because total costs and expenses grew a slower 15%, income from operations — what's left of revenue after the costs of running the business, before interest and tax — rose 24% to $758 million. That lifted operating margin from 19.8% to 21.0%.
Net income rose faster, up 27% to $816 million, and diluted earnings per share faster still, up 33% to $1.37. Neither of those faster growth rates came from the business getting better. Both came from two items below the operating line, detailed further down.
The numbers
Metric
Q2 2026
Q2 2025
YoY change
Revenue
$3,608M
$3,096M
+17% (+13% excluding currency)
Income from operations
$758M
$612M
+24%
Operating margin
21.0%
19.8%
+1.2pts
Adjusted EBITDA
$1,261M
$1,043M
+21%
Adjusted EBITDA margin
Read 0 community reports on Airbnb, or write your own.Write a report
Three terms worth defining once. Gross Booking Value is the total dollar value of everything booked on the platform in the quarter — what guests pay, including the host's earnings, cleaning fees and taxes, net of cancellations. Airbnb keeps only its fees out of that, so take rate (revenue divided by GBV) is the share of booking dollars the company actually books as its own revenue. Adjusted EBITDA is the company's own profit measure, which strips out tax, interest, depreciation and — most importantly here — stock-based compensation, which was $487 million in the quarter, up from $424 million. That single exclusion is why Adjusted EBITDA margin (35.0%) is 14 points higher than the operating margin (21.0%): the gap is mostly one real, recurring, non-cash pay expense.
What actually drove the quarter
Growth came more from volume than price, and more from newer markets than old ones. Nights and Seats Booked rose 10%, an acceleration from 9% in Q1 2026. ADR — the average price per night or seat booked — rose 5%, but only 4% excluding currency effects, and the 10-Q attributes part of even that increase to a mix shift rather than pure price: entire homes, and specifically listings with four or more bedrooms, grew fastest. Bookings measured on a bedroom basis (nights multiplied by bedroom count) grew over 12% against the 10% nights figure, meaning the average booking is getting physically larger, not just more expensive.
By region, the acceleration is real but concentrated in the small part of the business. Revenue by listing location:
Region
Q2 2026 revenue
Q2 2025 revenue
YoY change
Share of revenue
North America
$1,594M
$1,377M
+16%
44%
Europe, Middle East & Africa
$1,425M
$1,233M
+16%
39%
Latin America
$291M
$231M
+26%
8%
Asia Pacific
$298M
$255M
+17%
8%
Management highlighted roughly 20% nights growth in Latin America, high-teens in Asia Pacific, Brazil origin nights accelerating past 30%, and India origin nights up 60% with first-time bookers there more than doubling. Those are the fastest numbers in the letter, and together those two regions are 16% of revenue. The more consequential datapoint is that North America posted high-single-digit nights growth — its best in almost three years — and that net origin nights accelerated in the U.S., France, the UK and Australia. A 44%-of-revenue region reaccelerating matters more to the 2027 revenue line than a 60% growth rate in India does.
Currency did meaningful work in the headline. Revenue grew 17% as reported but 13% excluding exchange-rate moves, so roughly a quarter of the reported growth was the dollar weakening against other currencies rather than more or pricier bookings. In Latin America, the 9% ADR increase was "primarily driven by FX" — on a constant-currency basis it was 2%.
The expense mix is the most interesting part of the filing
Operating leverage this quarter was not broad-based. It came from one line shrinking in relative terms while another was deliberately opened up.
Expense line
Q2 2026
Q2 2025
YoY change
Cost of revenue
$633M
$544M
+16%
Operations and support
$361M
$332M
+9%
Product development
$672M
$610M
+10%
Sales and marketing
$875M
$691M
+27%
General and administrative
$309M
$307M
+1%
Total costs and expenses
$2,850M
$2,484M
+15%
Sales and marketing was the single largest increase, up $184 million, of which $132 million was marketing spend "driven by higher paid growth marketing initiatives in emerging markets and partnerships" and $48 million was payroll from higher headcount. This is the company buying the Latin America and Asia Pacific growth rates above, and it grew 10 points faster than revenue.
What paid for it was operations and support, up only 9% despite a 16% increase in booking volume. The 10-Q is specific: a $17 million decrease in third-party service provider costs "due to lower agent contact volume resulting from increased use of artificial intelligence in community support," partly offsetting $27 million more payroll, $10 million more in customer make-good payouts, and $7 million more in host liability insurance premiums. The shareholder letter puts a number on it — customer support cost per booking fell approximately 16% year over year, with nearly 45% of issues that start with the AI assistant now resolved without a human agent. That is the clearest instance in this filing of AI showing up as a measurable cost line rather than a narrative, and it is currently funding the marketing expansion rather than dropping to margin.
Cost of revenue's 16% increase was mostly mechanical: $68 million more in merchant fees on higher payment volume, $13 million more in chargebacks (on both higher GBV and a slightly worse chargeback rate), $12 million more in server costs. Product development's $62 million increase was entirely payroll. General and administrative was flat only because a $32 million payroll increase was almost exactly cancelled by a $28 million decrease in non-income taxes — a netting artifact, not cost discipline.
Why EPS grew 33% when operating income grew 24%
Two items, neither of them operational:
A $77 million tax benefit. The provision for income taxes fell 41% to $81 million, "primarily due to a $77 million benefit recorded in the current period related to recently published guidance impacting prior year taxes," partly offset by $9 million of additional tax on higher profits. The effective tax rate was 9.0% of pre-tax income, against 17.6% a year earlier. Management guides the full-year 2026 effective rate to the high teens, which marks this as a one-quarter discrete item. At last year's rate, net income would have been roughly $739 million — about 15% growth, in line with revenue rather than ahead of it.
A smaller share count. Diluted shares fell 4.6% to 597 million after $1.1 billion of Class A buybacks in the quarter ($2.1 billion in the first half). That converts 27% net income growth into 33% EPS growth without any change in the business.
Working the other way, and worth noting because it is now permanent: net interest income fell from $184 million to $146 million. In March 2026 Airbnb issued $2.5 billion of senior notes (tranches at 4.40% due 2029, 4.65% due 2031 and 5.25% due 2036) and used part of the proceeds to repay $2.0 billion of 0% convertible notes at maturity. Swapping zero-coupon convertible debt for coupon-bearing debt took interest expense from $6 million to $37 million, while interest income slipped to $183 million on lower rates. That roughly $38 million annualizing headwind will recur every quarter; the $77 million tax benefit will not.
Takeaway: The 17% revenue growth and 120 basis points of operating margin expansion are real, but the quarter's two most quotable numbers are not what they look like — 27% net income growth and 33% EPS growth shrink to roughly 15% once the one-off $77 million tax benefit and the 4.6% reduction in share count are removed, which is the same rate as revenue. The durable finding is in the expense mix: AI cut support costs per booking by about 16% and took $17 million out of outsourced agent spend, and management immediately redeployed that and more into a 27% increase in sales and marketing aimed at Brazil, India and other expansion markets. Airbnb is currently choosing growth over margin, and the margin gain happened anyway.
Cash generation is strong, but the quality of it is changing
Free cash flow of $1.253 billion rose 30%, a 35% margin on revenue, and trailing-twelve-month free cash flow reached $4.827 billion. But the first-half picture shows a developing drag. Operating cash flow for the six months was $2.978 billion against $2.764 billion a year earlier — up only 8% while revenue grew 17%.
The cause is disclosed plainly: unearned fees grew "at a rate less than the GBV growth rate," because of "increased guest adoption of our flexible payment options, which allow guests to pay closer to check-in dates rather than at time of booking." Reserve Now, Pay Later is the specific product. Unearned fees — money collected for stays that haven't happened yet — were $2.831 billion at June 30, 2026 versus $2.857 billion a year earlier: essentially flat while GBV grew 16%. Management says that absent Reserve Now, Pay Later, unearned fees would have grown.
This matters for two reasons. First, Airbnb's historically enormous free cash flow has partly been a float business — collecting guest money at booking and holding it until the stay. As more guests pay later, less float accumulates, and free cash flow converges toward actual earnings. Second, the same product is cited as a factor affecting the take rate, alongside currency: guests booking further in advance shifts the timing between when a booking is counted in GBV and when the stay generates revenue. Take rate came in at 13.2%, level with a year earlier, and management expects it to stay roughly flat year over year in Q3.
Separately, Airbnb is standardizing its fee structure — migrating most hosts from a split arrangement (a 3% host fee plus a separate guest service fee) to a single 15.5% host fee, with the remaining migration expected to finish this year. Hosts can reprice to keep the same net earnings. This is presented as transparency rather than monetization, and the flat take rate is consistent with that.
Newer lines, and one metric caveat
Experiences supply grew nearly 80% year over year, with seats booked accelerating both year over year and sequentially. Services expanded into grocery delivery, car rentals, airport pickups, luggage storage and resort passes. Boutique and independent hotels were added across more than 20 destinations; hotel nights are still a single-digit percentage of total nights but grew roughly three times as fast as the homes business, and about 35% of guests who first book a hotel on Airbnb come back to book a home. Guest travel insurance revenue, available in 12 countries, grew over 60%.
The caveat: the headline volume metric is "Nights and Seats Booked," which sums lodging nights with seats booked for experiences and services. As Experiences supply grows 80% and services scale, that metric increasingly blends two businesses with very different economics and price points, and its 10% growth is no longer a clean read on lodging demand. Airbnb does not break the two apart. ADR is now reported as GBV per night and seat booked, so the same blending affects the price metric.
Regulatory exposure is the largest unquantified risk
Spain's Ministry of Consumer Affairs proposed a fine of roughly €110 million ($129 million) in July 2025 over alleged non-compliance with short-term rental listing rules, then reduced it to about €65 million ($76 million) in September 2025. Airbnb disputes both the fine and the applicability of the rules, and has not accrued for it — the filing states a loss is "neither probable nor estimable." In May 2026, under a court order, Airbnb posted a €70 million ($80 million) surety bond to suspend enforcement pending resolution, with no fixed expiration. Posting a bond does not concede the claim, but it does mean enforcement was live enough to require security.
On non-income taxes (lodging, value-added, digital services and host withholding taxes across multiple jurisdictions), accrued obligations rose to $219 million at June 30, 2026 from $199 million at year-end 2025, with a further $240–260 million of reasonably possible exposure above what's accrued. The filing notes flatly that "global regulatory requirements and challenges affecting the Company's business continue to increase."
Guidance and trajectory
Management raised full-year 2026 guidance on both revenue and margin:
Guidance
Prior
Now
FY2026 revenue growth
—
at least mid-teens (raised)
FY2026 Adjusted EBITDA margin
—
at least 35.5% (raised)
Q3 2026 revenue
—
$4.69B–$4.77B, +15% to +17%
Q3 2026 GBV growth
—
mid-teens
Q3 2026 Nights and Seats Booked
—
low double-digit growth
Q3 2026 Adjusted EBITDA margin
—
down slightly vs. Q3 2025's 50%
FY2026 effective tax rate
—
high teens
Two things to read carefully in that Q3 guide. The 15–17% revenue growth range is stated as "inclusive of an approximate three percentage point FX tailwind after factoring in our hedging program" — so the underlying growth being guided is roughly 12–14%, a touch below Q2's 13% constant-currency result. And the Adjusted EBITDA margin is guided down year over year "due to timing of investments," against a Q3 2025 comparison of 50%. Third quarter is the peak travel season for North America and EMEA and carries by far the highest margin of the year; Q2's 35% is not the run rate in either direction.
My read: the operating story is more solid than last year, and the Q2 margin expansion is credible because it came from a disclosed, mechanical cost reduction in support rather than from a pause in spending. But the second half asks investors to accept lower margin for growth investment, at the same time that currency support is guided to shrink from roughly four points to three, and at the same time that the tax rate normalizes from 9% back toward the high teens. All three push reported earnings growth below revenue growth in Q3 and Q4, even if the business itself keeps improving. The number to watch next quarter is not revenue — it is whether nights growth holds at low double digits in North America and EMEA once the World Cup 2026 boost, which added more than 150,000 first-time-host listings across host cities, is no longer in the comparison. That event is a one-time supply and demand injection, and it lands in the third-quarter base.
The cash flow line deserves attention too. If unearned fees stay flat while GBV compounds in the mid-teens, operating cash flow will keep growing slower than revenue, and the gap between Airbnb's reported free cash flow and its economic earnings will close. That is not deterioration — it is the float unwinding as guests pay later — but anyone valuing the company on a trailing free-cash-flow multiple should expect that multiple to look less generous over the next few quarters.
Source: Airbnb, Inc. Form 10-Q for the quarterly period ended June 30, 2026, filed with the SEC on August 6, 2026 (CIK 0001559720), and the Q2 2026 shareholder letter filed the same day as Exhibit 99.1 to Form 8-K. All figures are as reported. Adjusted EBITDA and free cash flow are non-GAAP measures as defined and reconciled by the company.