Meta's Q3 Beats Forecasts, but Free Cash Flow Is the Number to Watch
Meta Platforms reported third-quarter 2026 revenue of $64.60 billion and diluted earnings per share of $3.83 on Wednesday, October 21, after the US market closed, according to the company's press release. The company had scheduled a conference call for 1:30 p.m. PT (4:30 p.m. ET) on the day of the announcement.
Revenue came in above the $61 billion to $64 billion range Meta had guided to for the quarter in its second-quarter release in July. Bloomberg analyst estimates, as Yahoo Finance reported on the day of the release, had called for EPS of $3.56 on revenue of $63.3 billion. Against that consensus, EPS was about 7.6 percent higher and revenue about 2.1 percent higher.
Sequentially, the picture is less tidy. Revenue rose from $60.80 billion in the second quarter, a gain of roughly 6.3 percent. Diluted EPS fell from $6.18, a decline of about 38 percent. The second-quarter release had reported revenue up 28 percent year over year. Meta's first-quarter 2026 release reported diluted EPS of $10.44, so per-share earnings have moved a long way in two quarters even as revenue kept climbing.
Part of that volatility is accounting. Meta's first-quarter release notes an $8.03 billion income charge recorded in the third quarter of 2025. Any year-over-year EPS comparison against that quarter will be distorted by it. Investors who want to strip out one-off items should check the reconciliation tables in the release, the schedules that walk from reported figures to adjusted ones, rather than rely on headline per-share growth.
In January, Bloomberg reported that improvements in the online advertising business were making it possible for Meta to spend hundreds of billions of dollars on AI over the next few years. Ahead of this report, Reuters said Meta's AI outlays and the adoption of its Muse personal assistant would be in focus. A Reuters Breakingviews column dated October 21 projected negative free cash flow of $8.5 billion for the third quarter, based on compiled estimates.
Free cash flow is operating cash flow minus capital expenditure, the spending on long-lived assets such as data centers. It shows whether the AI build-out is being funded from the business or from the balance sheet. A negative figure means the company spent more cash than its operations generated. For anyone sizing Meta's capacity to keep spending, it is the line that matters most.
Muse is the product thread running through the quarter. Meta announced Muse Spark in April, calling it its most powerful model to date and the engine for a smarter Meta AI assistant across WhatsApp, Instagram, Facebook, and Messenger. In July, the company said Meta AI, powered by Muse Spark 1.1, could make plans, connect to email and calendar apps, create slides, and handle tasks on users' behalf. In September, Meta introduced Muse as a personal AI agent that proactively helps people with their goals, and on September 29 it followed up with a newsroom post on Muse for small business.
The commercial question is whether agentic features, meaning AI that acts for users rather than only answering them, translate into engagement and ad revenue quickly enough to justify the capital being committed. Reuters' reporting suggests that was the question the market was asking before the print.
Meta announced a quarterly cash dividend on September 10, 2026.
In my view, the dividend matters mainly as a signal that the company is maintaining its payout while it leans into capital spending. It says little about whether that spending earns a return.
Bloomberg reported on July 29 that Meta shares fell after the company defended its AI bets to skeptical investors and gave a revenue forecast that disappointed. A Bloomberg opinion column dated October 21 framed the report around investors looking to Mark Zuckerberg for the payoff from Muse.
That framing is the right one to hold onto. A beat on revenue and EPS against a consensus set in the weeks before the print tells you the top line is running ahead of expectations. It does not tell you what the spending earns.
In my view, professional investors should anchor the read on three numbers: the quarter's revenue against the guided range, the cash generated against the capital committed, and the expense outlook the company updates on the call. Headline EPS, distorted by prior-year charges and by swings in tax and one-off lines, is the least reliable of the group. The $3.83 is a real result. Whether it starts a trend or is a single quarter in a heavy investment cycle will take more than one release to settle.
