Meta Q2 2026: revenue up 28%, free cash flow down 91% to $784M

Meta’s second-quarter 2026 revenue was $60.80 billion, up 28% year over year and near the top of the $58–61 billion range the company guided to in its April 29 filing, while diluted EPS fell 13% to $6.18. That split is the quarter: the advertising business sold more ads at higher prices, and the profit line went backwards because of $3.58 billion of one-off charges and a cost base growing almost twice as fast as revenue.

The more consequential line is further down the release. Capital expenditures, including principal payments on finance leases, were $31.08 billion, and free cash flow fell to $784 million from $8.55 billion a year earlier, a 91% drop (our arithmetic), below the $1.35 billion Meta paid in dividends. Meta (NASDAQ: META) also narrowed its 2026 capex outlook upward, to $130–145 billion. Microsoft reported the same evening for the same calendar quarter; the two companies’ numbers are compared in our Microsoft Q4 FY2026 analysis.

Meta Q2 2026 results at a glance

	Q2 2026	Q1 2026	Q2 2025
Revenue	$60,801M	$56,311M	$47,516M
Revenue growth, YoY	28%	33%	22%
Total costs and expenses	$42,026M	$33,439M	$27,075M
Income from operations	$18,775M	$22,872M	$20,441M
Operating margin	31%	41%	43%
Net income	$15,848M	$26,773M	$18,337M
Diluted EPS	$6.18	$10.44	$7.14
Capex incl. finance-lease principal	$31.08B	$19.84B	$17.01B
Free cash flow	$784M	$12,386M	$8,549M
Family daily active people	3.60B	3.56B	3.48B
Headcount	75,472	77,986	75,945

Q1 2026 net income and EPS include an $8.03 billion income tax benefit tied to Treasury Notice 2026-7, so that column is not a clean comparison for the bottom line. Free cash flow is Meta’s own non-GAAP measure: cash from operations minus purchases of property and equipment minus principal payments on finance leases.

By segment, Family of Apps (Facebook, Instagram, Messenger, WhatsApp) produced $60,370 million of revenue and $23,394 million of operating income, down from $24,971 million a year earlier. Reality Labs produced $431 million of revenue and a $4,619 million operating loss. The effective tax rate was 16%, against 11% a year ago.

Did Meta meet its own guidance in Q2 2026?

Yes on revenue: $60.80 billion sits inside the $58–61 billion range from the April 29 filing, $1.3 billion above the $59.5 billion midpoint (our arithmetic). That guide assumed foreign currency would be an approximately 2% tailwind; the release’s constant-currency table puts the actual tailwind at about one point, with revenue up 27% excluding currency effects. So the result near the top of the range came with less help from currency than Meta had assumed.

On expenses, the April filing guided full-year 2026 total expenses to $162–169 billion. The July filing raised the lower end to $165–169 billion “to incorporate the $2.4 billion charges related to legal proceedings recognized in the second quarter,” and repeated that Meta expects 2026 operating income above 2025’s. Meta does not guide quarterly EPS, so there is no company figure to hold the $6.18 against other than the year-ago $7.14.

The growth rate itself slowed. Revenue grew 33% in Q1 2026 and 28% in Q2; ad impressions grew 14% against 19% in Q1, while average price per ad grew 12% in both quarters. The comparison with the 22% of Q2 2025 makes this quarter look like an acceleration; the comparison with the quarter just before it does not.

How much of the EPS decline was one-off charges?

Most of it. Costs and expenses include $2.40 billion of charges related to legal proceedings, booked in general and administrative expense, and $1.18 billion of severance tied to the May 2026 reduction of approximately 8,000 employees. Neither the release nor the 10-Q says which proceedings the legal charge relates to; the 10-Q’s contingencies note describes youth-related actions among other open matters, and the CFO outlook warns that youth-related trials “may ultimately result in a material loss.”

Together the two items are $3.58 billion before tax. Applying the quarter’s 16% effective tax rate as a rough proxy gives about $3.0 billion after tax, or about $1.17 per diluted share on 2,566 million shares, which would put EPS near $7.35 without them (our arithmetic). That is an approximation, not a company figure: Meta publishes no adjusted EPS, and if part of the legal charge is not tax-deductible, the after-tax effect is larger.

Even without the charges, costs grew far faster than revenue. Research and development rose 67% to $21,656 million, which the 10-Q attributes primarily to employee compensation, infrastructure and third-party cloud services, and third-party AI token costs; cost of revenue rose 33%, which it attributes to infrastructure expenses related to data centers, technical infrastructure and third-party cloud services. Depreciation and amortization was $6,356 million, up from $4,342 million, and depreciation on servers and network assets alone was $4.62 billion against $3.12 billion a year earlier.

What did Meta guide for Q3 and the rest of 2026?

Meta guided third-quarter revenue to $61–64 billion, assuming an approximately 1% currency headwind. Against Q3 2025 revenue of $51,242 million, that range implies growth of about 19% at the low end, 22% at the midpoint and 25% at the top (our arithmetic). Every point in the range is slower than the 28% just reported.

The capital outlook moved the other way. Meta now expects 2026 capex, including principal payments on finance leases, of $130–145 billion, “narrowed from our prior outlook of $125–145 billion.” The ceiling did not change; the floor rose by $5 billion. Against full-year 2025 capex of $72.22 billion, the range is roughly 80% to 100% higher (our arithmetic). The 10-Q backs the floor with contracts: $349.31 billion of non-cancelable contractual commitments as of June 30, mostly third-party cloud capacity, servers, network infrastructure and data centers, with about $53.52 billion due in the rest of 2026 and $81.65 billion in 2027.

The tax outlook for the remaining quarters of 2026 rose to 15–17%, from 13–16%.

Why did free cash flow fall 91%?

Because capex grew faster than operating cash flow; the 91% decline is our arithmetic on the release’s $784 million and $8,549 million. Cash from operations was $31,862 million, up from $25,561 million, but purchases of property and equipment were $30,116 million and finance-lease principal $962 million, leaving $784 million. In Q2 alone, capex equalled about 51% of revenue (our arithmetic).

That $784 million covered about 58% of the $1,353 million of dividends and dividend equivalents paid (our arithmetic). Meta made no share repurchases in the quarter or the first half, per the 10-Q, with $25.03 billion still authorized; in Q2 2025 it had repurchased $9.76 billion of Class A stock. The gap was funded by debt: the 10-Q reports $24.91 billion of net proceeds from notes issued in May 2026, and long-term debt stood at $83.66 billion at June 30 against $58.74 billion at the end of 2025. Cash, cash equivalents and marketable securities were $90.26 billion, so cash exceeded long-term debt by about $6.6 billion (our arithmetic, before leases).

What still improved in the core business?

The ad engine grew on both of its levers. Ad impressions rose 14% and average price per ad rose 12%; 1.14 × 1.12 is about 1.28, close to the 27% growth in advertising revenue to $59,363 million (our arithmetic; the two metrics are company figures, the product is ours). Growth from both volume and price is a stronger configuration than growth from either alone.

Engagement kept growing: family daily active people averaged 3.60 billion in June 2026, up 3%, or about 120 million people (our arithmetic). Headcount fell 1% to 75,472, and the release says that figure still includes approximately 8,000 people affected by the May reduction, most of whom will be off the count by the end of Q3.

Reality Labs lost $4,619 million on $431 million of revenue, about $10.70 of loss per dollar of sales (our arithmetic), slightly more than its $4,530 million loss a year ago. Next to quarterly capex 6.7 times that loss (our arithmetic), it is no longer the line that decides Meta’s spending story.

What to watch in Q3 2026

• Revenue against the $61–64 billion range. The release gives the range; where the print lands inside it shows whether the slowdown from 33% to 28% continues.
• Capex against $130–145 billion. First-half capex was $50.92 billion (our arithmetic: $19.84 billion plus $31.08 billion), so the floor implies at least about $79 billion in the second half.
• Free cash flow against the dividend. A second quarter below the $1.35 billion payout would make debt the regular funding source for shareholder returns.
• Depreciation growth. Server depreciation rose about 48% year over year (our arithmetic); its pace against revenue growth is where the 2026 spending reaches the income statement.
• Legal matters. The CFO outlook flags youth-related trials in the US that “may ultimately result in a material loss.”

Quick answers

Did Meta beat its revenue guidance in Q2 2026?

Revenue of $60.80 billion came in near the top of the $58–61 billion range Meta guided to on April 29, 2026, and rose 28% year over year. Diluted EPS fell 13% to $6.18, and Meta gives no EPS guidance.

Why did Meta’s earnings fall in Q2 2026?

Costs and expenses rose 55%, including $2.40 billion of charges related to legal proceedings and $1.18 billion of severance from the May 2026 headcount reduction, plus higher infrastructure, depreciation and compensation costs. Net income fell 14% to $15.85 billion.

What is Meta’s 2026 capex guidance?

$130–145 billion including principal payments on finance leases, narrowed from $125–145 billion in April. Q2 capex was $31.08 billion.

---

Sources and method. Company figures — the income statement, segment results, balance sheet, cash flows, free-cash-flow reconciliation, ad metrics, DAP, headcount, the legal and severance charges, and the CFO outlook — are Meta’s own disclosures in Exhibit 99.1 to the Form 8-K filed 2026-07-29 (accession 0001628280-26-050596) for the quarter ended 2026-06-30, read directly from SEC EDGAR and linked above; the filing was accepted at 16:03 ET. Depreciation by asset class, the absence of share repurchases and the remaining authorization, the May 2026 notes, contractual commitments and the expense-line explanations come from the Form 10-Q filed 2026-07-30 (accession 0001628280-26-050705). The Q2 2026 revenue guidance and prior expense and capex outlooks come from Exhibit 99.1 to the Form 8-K filed 2026-04-29; year-ago, Q3 2025 and full-year 2025 comparatives come from Meta’s earlier 8-K exhibits, all linked above.

No consensus estimate is included, so this piece makes no claim that the quarter beat or missed expectations; the only comparisons are with Meta’s own guidance and with prior periods. This piece makes no claim about the share-price reaction.

Our arithmetic, labelled as such: the 91% decline in free cash flow; the $1.3 billion gap to the guidance midpoint; the ~$1.17 per-share effect of the charges and the ~$7.35 EPS without them (a 16% tax proxy on 2,566 million diluted shares); the implied 19%, 22% and 25% growth in the Q3 outlook; the ~80% to ~100% increase implied by the 2026 capex range; capex at ~51% of revenue; free cash flow covering ~58% of dividends; the ~$6.6 billion of cash over long-term debt; the 1.14 × 1.12 ad reconciliation; the ~120 million DAP increase; the ~$10.70 Reality Labs loss per revenue dollar and the 6.7× capex multiple; first-half capex of $50.92 billion and the ~$79 billion second-half floor; and the ~48% growth in server depreciation. These are approximations built on the disclosed figures, not reported numbers. The interpretation — that the charges explain most of the EPS decline, that the Q3 outlook and the capex floor matter more than the charges, and that free cash flow below the dividend is the quarter’s most consequential figure — is our analysis, not a company statement, and is not a forecast.

Related: Microsoft Q4 FY2026, reported the same evening for the same calendar quarter, and the next day’s Amazon Q2 2026 and Apple Q3 FY2026.

Tradune shows what you actually hold across every account, and whether a position has beaten a plain index fund since you bought it. Educational analysis, not investment advice.