Meta Q3 2026 Earnings: Revenue Exceeds Expectations, Free Cash Flow Turns Negative Amid AI Investments

Original | Odaily Planet Daily (@OdailyChina)

Author | Azuma (@azuma_eth)

Amid the shadow of an "AI confidence crisis," Zuckerberg was the first to reassure the market.

Over the past few weeks, skepticism has grown regarding major tech giants' heavy investments in AI and their prolonged inability to demonstrate tangible business returns, leading to a noticeable correction in chip stocks and Nasdaq tech shares. As the first trillion-dollar tech giant to report earnings in this earnings season, Meta’s Q3 2026 financial results, released after U.S. markets closed on October 21 Eastern Time, became the first major test of AI’s commercial viability.

Financial results showed that Meta's third-quarter revenue reached $64.6 billion, a 26% year-over-year increase, surpassing analysts' expectations of $63.3 billion. The company forecasts fourth-quarter revenue between $73 billion and $76 billion, with the midpoint also exceeding market expectations. In response, the market has initially welcomed these results—Meta’s stock rose as much as 6% in after-hours trading following the stronger-than-expected quarterly revenue and fourth-quarter guidance, though it has since retreated to a 2% gain.

However, a closer look at the financial statements reveals a strong sense of dissonance. On one side, there’s a quarterly revenue of $64.6 billion, steadily growing advertising business, and the AI assistant Muse, which attracted over ten million users in North America within just one and a half months of launch. On the other side, the company’s free cash flow turned negative for the first time since its IPO, with nearly $10 billion in legal settlement expenses recognized for the quarter, full-year capital expenditure guidance raised further, and Meta announcing a new Class C common stock plan to reserve capacity for future financing and equity incentives.

Putting these factors together paints a clear picture of Meta’s current position—its existing businesses continue to generate ample revenue and cash flow, but Zuckerberg is clearly unsatisfied; he is redirecting more resources toward AI and securing funding in advance for what may be a longer-term competition.

First, let’s look at the income statement, which is most likely to be misunderstood.

Meta's total costs and expenses in Q3 increased by 68% year-over-year to $51.61 billion, significantly outpacing revenue growth of 26%; GAAP operating profit declined by 37% year-over-year to $12.99 billion; the operating margin also dropped from 40% in the same period last year to 20%.

Looking at these numbers alone, it’s hard to call this a strong income statement. However, the decline in profits was primarily driven by a one-time expense—the August settlement with attorneys general from all 50 U.S. states and territories regarding lawsuits related to adolescent mental health and social media addiction. According to the sources cited, this arrangement involves potential payments of up to approximately $18.59 billion over the next decade. Meta recorded a one-time legal settlement expense of $9.89 billion in the third quarter, classified under general and administrative expenses (G&A).

Excluding this one-time expense, Meta's operating profit for the quarter was approximately $22.88 billion, and diluted earnings per share were approximately $7.40, significantly higher than the reported $3.83 and exceeding Wall Street’s prior expectations of $6.08 to $6.54.

The solid foundation of this genuine profitability remains Meta’s untouchable core advertising business. In the third quarter, Meta’s total revenue reached $64.6 billion (a 26% year-over-year increase), surpassing the market expectation of $63.3 billion; of this, advertising revenue from the Family of Apps (FoA) contributed $63.06 billion, also up 26% year-over-year. What further reassured investors was the internal structure of advertising growth—both volume and price increased.

• Impressions: Total app advertising impressions increased by 10% year-over-year, with daily active persons (DAP) reaching 3.64 billion (up 3% year-over-year);
• Average price per ad: The average price per ad increased by 14% year-over-year.

In today’s macroeconomic environment filled with uncertainty, a 14% increase in ad pricing is highly significant. This directly validates Meta’s two-year strategy of pouring massive computational resources into recommendation algorithms and automated advertising tools—such as Advantage+, which now generates annualized revenue exceeding $75 billion—creating a positive feedback loop: AI has substantially improved ad conversion rates and advertisers’ ROI, enabling Meta to maintain strong pricing power even as social traffic reaches its peak.

Meanwhile, Meta’s Q4 revenue guidance of $73 billion to $76 billion (a midpoint of $74.5 billion, representing approximately 24% year-over-year growth) once again comfortably exceeded Wall Street’s consensus expectation of around $73.7 billion, demonstrating that advertising demand during the year-end holiday shopping season remains extremely strong.

If the advertising business demonstrated AI's power in improving efficiency within existing operations, then Muse, the AI personal assistant launched in September, has given Wall Street its first glimpse of Meta's potential to monetize new, AI-native applications.

During the earnings call, Zuckerberg could not hide his excitement about Muse and Muse Wonder, the cutting-edge foundational model released last week. Based on the key operational data disclosed, Muse is helping Meta transition its narrative from an "AI money-burner" to an "AI revenue-generator."

• User Growth and Engagement: In just one and a half months since launch, over 10 million users in the United States and Canada have created Muse personal agents, with approximately 25% of these users converting into daily active users (DAUs). Early retention curves have become increasingly stable as usage duration increases.
• Break the myth of “chatting without earning”: Unlike many large language model applications that remain limited to Q&A, shopping has become Muse’s largest use case. Zuckerberg revealed that over a quarter of Muse users have made actual purchases on the platform, with an average order value (AOV) exceeding $200. Combined with the monthly subscription tiers of $20 and $100, which officially launched in mid-October, Muse now holds dual revenue potential through “subscription + e-commerce affiliate commissions.”
• Integrated hardware and software with B-end expansion: Starting next month, Muse will officially launch on Meta’s smart glasses, becoming the core of its hardware business model; on the enterprise side (B2B), CFO Susan Li has clearly stated that developers will be charged for access to the foundational model and Muse Code programming agents, with plans to gradually build a “Muse Business Agent Network” serving small and medium-sized enterprises.

Of course, concerns remain. Dozens of external websites, including Amazon, have blocked access to Muse due to data security and business competition considerations, somewhat limiting its seamless cross-platform price comparison and shopping experience. However, Zuckerberg’s cost roadmap remains highly compelling—Muse’s computational cost efficiency is expected to increase tenfold within the next year and become a highly profitable business within a few years.

Turn to the back of the earnings report, and the cost Zuckerberg has paid for this AI gamble is equally staggering.

To gain an edge in the compute arms race against OpenAI, Anthropic, and Google, Meta’s capital expenditures for the third quarter alone—including principal payments on lease financing—reached $34.91 billion. For the first nine months of the year, spending on property and equipment totaled $83.06 billion, nearly doubling compared to $48.31 billion in the same period last year.

This alarming cash consumption directly led to a historic financial turning point—despite Meta’s operating cash flow of $32.79 billion in Q3, its free cash flow fell to -$2.12 billion after capital expenditures (compared to $10.63 billion in the same period last year), marking the first time since its IPO that Meta has recorded negative free cash flow.

Even so, Zuckerberg has no intention of slowing down. Meta raised its full-year 2026 capital expenditure guidance floor from $130 billion to $135 billion (narrowing the range to $135 billion–$145 billion) and explicitly stated it will continue “aggressive investment” in 2027.

With operating cash flow under pressure and long-term debt rising from $58.74 billion at the end of last year to $80.92 billion, where will the hundreds of billions of dollars needed for future computing power come from? Meta provided the answer in its third-quarter earnings report—new Class C common shares.

According to the company’s announced plan, Meta will distribute Class C shares to existing Class A and Class B common shareholders on a 1:1 basis, with the arrangement expected to take effect on January 19, 2027. Since Class C shares carry no voting rights, this plan separates economic entitlement from corporate control. For Meta, this arrangement holds at least three key implications.

• First, it provides more options for future equity financing and mergers and acquisitions. As investment in AI infrastructure continues to rise, Meta must consider not only how to increase capital expenditures but also how to balance cash, debt, and equity financing. Introducing a new class of non-voting shares offers a tool for future fundraising and equity incentives.
• Second, this helps maintain the existing control structure. Meta’s Class A shares carry one vote per share, while Class B shares carry ten votes per share. Creating non-voting Class C shares allows the company to expand its equity instruments to some extent without altering the current voting rights arrangement for existing shareholders. This is particularly important for Zuckerberg, who seeks to retain firm control over the company.
• In addition, to align with the new share plan, the board entered into a Founder Alignment Agreement with Zuckerberg. Previously, due to the potential billions of dollars in tax liability from the proposed California Wealth Tax (Proposition 40), there were market rumors that Zuckerberg might leave California. However, through this capital structure restructuring and the design of sale provisions, Zuckerberg has now committed to remaining in California and continuing his leadership for the next eight years.

It is important to emphasize that creating Class C shares does not mean Meta has decided to raise its next round of AI funding through a large-scale share issuance, nor does it directly indicate an imminent funding shortfall. Rather, it is more like proactively preparing a financing tool to preserve greater flexibility for the future as investment scales continue to grow.

This initiative also signals another point—that the AI race may not just be a short-term surge in investment over a few quarters, but a long-term battle requiring sustained effort over many years and upfront planning of capital structure.

To free up more resources for AI, Meta is also restructuring its internal operations and organization.

First, the metaverse division (Reality Labs), which has consistently been a drag on operations, will officially undergo a strategic downsizing. In the third quarter, Reality Labs generated $443 million in revenue but incurred an operating loss of $4.13 billion (bringing the cumulative loss for the first nine months to $13.17 billion). Unlike previous vague statements about losses "peaking," the CFO provided an unusually clear "timeline for reducing losses" in this earnings outlook: Reality Labs' full-year loss is expected to narrow to $18–19 billion in 2026, and further decline to approximately $15 billion in 2027. The billions of dollars in savings are expected to be reallocated to AI initiatives centered around Meta’s Superintelligence Lab (MSL).

Second is the “blood transfusion” of the organizational structure. As of September 30, 2026, Meta’s total employee count had dropped to 66,775, a significant year-over-year reduction of 15% (and approximately 12% quarter-over-quarter). Yet, despite this major workforce reduction, Meta’s R&D expenses for the first three quarters surged from $40.24 billion in the same period last year to $59.41 billion. Behind this “shrinking headcount, exploding R&D” disparity lies the large-scale elimination of general roles, while resources are being concentrated with unprecedented intensity toward exorbitant AI computing power and top-tier large model researchers.

Overall, Meta won its first defensive battle against Wall Street’s “AI confidence crisis” with a strong third-quarter earnings report, optimistic guidance, and the impressive launch of Muse, which surpassed ten million users. The post-market stock price increase demonstrates that the market remains willing to tolerate short-term negative free cash flow, as long as its core advertising cash machine continues to run and new AI products show tangible conversion potential.

But this does not mean the alarm has been completely lifted. With the full-year expense guidance raised to $175 billion to $179 billion and regulatory headwinds far from cleared, Meta has little room for error ahead.

When hundreds of billions of dollars in capital expenditures translate into massive financial depreciation by 2027, will the incremental revenue from Muse and AI advertising truly outpace the exponentially growing costs of computing power? Zuckerberg’s high-stakes gamble will eventually have to answer this ultimate question.