Meta Platforms and Microsoft diverge on AI profitability as Big Tech capex soars past $600B

Two of the largest companies on Earth just reported earnings within hours of each other, and the market’s reaction could not have been more different. Microsoft posted roughly $90 billion in quarterly revenue, an 18% year-over-year jump, and was rewarded with an 8% stock price surge. Meta reported $60.8 billion in revenue, a faster 28% growth rate, and got punished with a 10% decline.

The numbers behind the divergence

Microsoft’s Q2 2026 results were powered largely by Azure, the company’s cloud computing division. Its AI business within Azure hit a $37 billion annual run rate, up 123% year-over-year.

Meta told a different story. While its top line grew faster than Microsoft’s in percentage terms, profits fell 14% to $15.85 billion. The culprit was infrastructure spending, with Meta’s estimated 2026 capital expenditure landing somewhere between $115 billion and $135 billion, potentially even higher.

The capex arms race is just getting started

Projected AI-related capital expenditures across major tech companies are expected to reach $635 billion to $665 billion in 2026, up from approximately $381 billion in 2025. That’s roughly a 70% increase in a single year.

Microsoft is channeling spending through Azure, where enterprise customers pay for AI compute on a subscription basis. Meta, by contrast, is building infrastructure primarily to serve its own platforms, with its AI models powering recommendation algorithms, advertising targeting, and its family of apps.

Why crypto markets should be paying attention

AI-linked digital assets have become one of the more prominent narratives in crypto for 2026, with tokens tied to decentralized compute and AI utility seeing sustained interest. When Big Tech’s AI capex jumps from $381 billion to potentially $665 billion in a single year, the demand for compute resources is genuinely enormous, and not all of it can be served by three or four hyperscalers.

Token utility models in the decentralized compute space work somewhat like this: providers stake tokens to offer GPU capacity, users pay in tokens for access, and the network coordinates matching supply with demand.

What investors should watch next

Microsoft’s Azure AI business growing at 123% annually suggests the centralized players are scaling quickly. But a projected capex figure north of $635 billion also suggests demand is outpacing even their ability to build.

Meta’s situation adds another dimension. If its profit compression continues, it may eventually look to external compute providers rather than building everything in-house. A company spending $115 billion or more annually on infrastructure has strong incentives to find cheaper alternatives wherever they exist.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.