Meta Platforms delivered a clean beat on the numbers investors were watching. Earnings of $7.40 per share, excluding the attorneys general settlement charge, topped the $6.52 consensus by 11.3% and cleared the $6.65 Earnings Whisper number. Revenue of $64.60 billion beat the $63.38 billion consensus by 1.9% and landed above the top of management's $61–64 billion guidance range. The growth profile is far less flattering. Revenue grew 26.1% year over year, but adjusted earnings grew only 2.1%. That gap is the central tension of the quarter: Meta's advertising engine is still compounding, while the cost of its AI ambitions is absorbing nearly all of the incremental profit.

The quality of the beat deserves scrutiny. GAAP EPS was $3.83 after a $9.89 billion charge tied to the teen mental health settlement with 52 attorneys general, which provides for payments of up to about $18.59 billion. GAAP operating income fell 37%, and operating margin dropped to 20% from 40% a year ago, 31% in Q2 and 41% in Q1. Even excluding the charge, operating income of $22.88 billion rose only about 11%, and the adjusted margin remains well below last year's level. Research and development jumped to $20.05 billion from $15.14 billion. Capital expenditures, including finance lease principal, reached $34.91 billion, which pushed free cash flow to negative $2.12 billion. That compares with $784 million in Q2 and $12.4 billion in Q1. Meta made no share repurchases during the quarter. The call analysis puts debt at $83.7 billion against $82.4 billion in cash. The balance sheet shows long-term debt of $80.9 billion, so the debt figure depends on what is included, but either way the net cash cushion is largely gone.

Beneath the headline, the monetization mix is shifting. Ad impression growth slowed for a third straight quarter, from 19% in Q1 to 14% in Q2 to 10% in Q3, while revenue growth eased from 33% to 28% to 26%. Pricing filled much of the gap. Average price per ad rose 14%, up from 12% in each of the prior two quarters, which supports management's argument that AI-driven ranking and recommendations are lifting ad effectiveness. Partnership Ads passed a $15 billion run rate, up from $10 billion in Q1. Daily active people reached 3.64 billion, and Instagram and Facebook Reels time spent hit records. Headcount fell 15% year over year to 66,775 after the May reduction, a cost lever that only partly offsets compute spending.

The call preserved a credible growth story and was markedly more confident in tone. Zuckerberg said the Muse personal agent has "clear product market fit." More than 10 million U.S. and Canadian users set it up in roughly six weeks, which he said is over twice Threads' starting level. Muse Wonder shipped as Meta's first frontier model. Management also laid out several new revenue channels:

• an enterprise platform spanning a model API, Muse Code and business agents;
• an agent-to-agent commerce protocol with Walmart, Stripe, Shopify and Sierra;
• MetaOne, which has passed 15 million subscriptions and trials.

Management expects strong revenue growth throughout 2027, with new AI businesses beginning to contribute. It also expects Reality Labs losses to narrow to about $15 billion next year, down from $18–19 billion in 2026. Bears have legitimate ammunition, however. Muse is compute-intensive and not yet profitable. Full-year expense guidance rose to $175–179 billion, roughly the size of the charge. The capex floor moved up again to $135 billion. More youth-related trials are scheduled in the coming months. The commitment that 2026 operating income will exceed 2025 is also a point of disagreement. The release reiterates it only on an ex-charge basis, and the call analysis flags that the pledge was not repeated on the call.

The new non-voting Class C stock changes the setup. It will be distributed as a one-for-one dividend that functions like a two-for-one split. The board explicitly framed it as a tool for capital raising, compensation and acquisitions, alongside preserving founder control. Paired with negative free cash flow and rising debt, it signals that Meta is preparing to finance the AI buildout partly with equity, which raises real dilution questions.

Investor sentiment improved from -0.057 to roughly neutral at -0.001, a meaningful thaw but not enthusiasm. The stock entered the report up 32% from the $526 open after the July release and 9.8% above its 200-day moving average. It was also 11% below the $779.82 52-week high set on September 24, so some of the pre-report run had already been given back. Earnings Whispers trend readings were mostly positive heading in, with price, sequential growth and momentum all improving.

The bottom line is that Meta beat comfortably and showed real product traction with Muse, but the earnings leverage that once defined the story has disappeared. Adjusted EPS barely grew despite 26% revenue growth. Free cash flow turned negative, and a new share class opens the door to equity funding. Investors now have to underwrite AI monetization that is promising but still largely unproven in the numbers.