Meta just posted the kind of earnings report that makes investors reconsider everything. Revenue grew 28% — and nobody cared. Here’s what the numbers tell us about where Meta is headed, and why strong growth can’t hide a cash crisis.
The Numbers That Spooked Wall Street
Meta reported Q2 2026 revenue of $60.8 billion, up 28% year-over-year. That beat analyst expectations of $60.2 billion. Advertising revenue alone reached $59.3 billion. By most standards, that’s a great quarter.
But earnings told a different story. Diluted EPS came in at $6.18, missing Wall Street’s $7.22 consensus by over 14%. Total expenses surged 55% to $42 billion, driven by infrastructure costs, AI-related hiring, $2.4 billion in legal charges, and $1.18 billion in severance tied to Meta’s May 2026 layoffs. Operating margin compressed from 43% to 31%.
Then came the number that really matters: free cash flow. Meta generated $31.9 billion in operating cash flow — and then spent almost all of it. Capital expenditures hit $31.1 billion, plus another $962 million in finance lease payments. That left just $784 million in free cash flow, a 91% collapse from the $10.9 billion Meta generated in Q2 2024.
Shares dropped nearly 10% in after-hours trading, falling to around $529 from a $585 close.
Why Revenue Growth Can’t Save You When Costs Double
Here’s the part most coverage misses. Meta isn’t struggling because its business is shrinking. It’s struggling because it’s spending money faster than it’s making it.
In H1 2026, Meta invested $50.9 billion in capital expenditures — compared to $30.7 billion in the same period of 2025. That’s a 66% jump. Meanwhile, the company narrowed its full-year capex guidance to $130–$145 billion, raising the floor by $5 billion. Full-year expenses are now projected at $165–$169 billion.
For context, as we covered when Big Tech’s combined AI capex hit $725 billion, this isn’t just a Meta problem. Google’s parent Alphabet reported its first-ever negative free cash flow of $5.9 billion last week after spending $44.9 billion in one quarter on AI infrastructure. The entire industry is burning cash at a historic rate.
But Meta has a unique problem the others don’t.
Meta’s Missing Safety Net
Google has Cloud, which grew 82% in Q2 to $24.8 billion. Amazon has AWS. Microsoft has Azure. All three hyperscalers can point to a booming cloud business that directly monetizes their AI infrastructure spending.
Meta has ads. That’s it.
CEO Mark Zuckerberg addressed this gap on the earnings call: “We’re getting a lot of offers for compute at a significant premium over what we paid for it.” That’s an acknowledgment, not a revenue line. Meta is the only major hyperscaler without a thriving cloud business — and it’s spending more on AI compute than ever.
The nascent “Meta Compute” initiative, which we explored when Meta’s cloud push triggered a $200 billion chip selloff in early July, is still in its infancy. There’s no revenue contribution yet. The company is essentially asking investors to trust that $130+ billion in annual spending will eventually create an entirely new business segment — while also funding Llama model training, personal AI agents, and a Reality Labs division that lost another $4.62 billion this quarter.
The BlackRock Move and What It Signals
One day before earnings, Meta announced a $14 billion data center venture with BlackRock in El Paso, Texas. BlackRock-managed funds will own 80%; Meta keeps 20% and leases the capacity back.
This is a cash flow strategy, not just a real estate deal. By offloading ownership to an outside investor, Meta accesses computing capacity without parking $14 billion on its own balance sheet. CFO Susan Li referenced the deal as part of a plan to “attract capital from a wide range of markets.”
Translation: Meta knows it can’t fund this buildout alone.
The $88 Billion Question
Reality Labs posted another $4.62 billion operating loss on just $431 million in revenue. Cumulative losses have now surpassed $88 billion since the division was created. Zuckerberg has said 2026 will be the peak year, with reductions starting in 2027 as focus shifts toward wearable AI. We’ll see.
Meta guided Q3 revenue to $61–$64 billion, with a midpoint of $62.5 billion — about 1% below analyst expectations.
The market isn’t buying the patience argument anymore. As a recent analysis from Saxo Markets noted, the risk across Big Tech AI spending is skewed toward further increases, and investors in mid-2026 are rotating away from hyperscalers and into chip suppliers who are actually profiting from this buildout.
For Meta specifically, the calculus is stark: the company needs Meta Compute to start generating real revenue, and soon. Because $784 million in quarterly free cash flow, for a company spending over $30 billion a quarter on capex, is not a sustainable trajectory. It’s a runway — and the question is how long it is.
FAQs
What is Meta Compute and how does it work?
Meta Compute is an internal cloud infrastructure project designed to sell access to Meta’s excess AI computing power and models to outside businesses. It’s led by Meta’s head of infrastructure and Superintelligence Labs leader Daniel Gross, and was first reported by Bloomberg in July 2026.
How does Meta’s AI spending compare to other Big Tech companies?
In 2026, the five largest U.S. hyperscalers — Microsoft, Alphabet, Amazon, Meta, and Oracle — committed roughly $725 billion in combined AI capex. Alphabet’s Q2 capex hit $44.9 billion; Meta’s reached $31.1 billion. Both saw free cash flow collapse as a result.
What happened to Alphabet’s free cash flow in Q2 2026?
Alphabet posted negative free cash flow of $5.9 billion in Q2 2026 — the first time since its 2004 IPO. Capital expenditures of $44.9 billion exceeded the $39.1 billion in operating cash flow, driven by massive AI data center construction and chip procurement.
Why did Meta lay off 8,000 employees in 2026?
Meta cut approximately 8,000 jobs in May 2026, targeting coordination, testing, content operations, and middle management roles. The company cited AI’s growing ability to automate these functions, while redirecting savings toward AI infrastructure investment.
How are AI chip stocks performing vs. Big Tech in 2026?
AI chip suppliers are outperforming the hyperscalers that buy from them. The iShares Semiconductor ETF (SOXX) gained 108% in H1 2026, while the Magnificent Seven basket fell 6.6%. Investors are rotating toward companies profiting from the AI buildout rather than those funding it.
The post Meta Drops 10% As They Burned Through Nearly All Its Cash in One Quarter appeared first on Memeburn.