Meta Vs. Coreweave: Despite 2032 Partnership, ‘Meta Compute’ Remains a Massive Threat to Coreweave

Meta Platforms (NASDAQ: META | META Price Prediction) and CoreWeave (NASDAQ: CRWV) just closed the books on their Q1 2026 reports with sharply divergent financial profiles. Meta printed $56.31 billion in revenue while raising capex guidance to $125 to $145 billion. CoreWeave grew sales 111.7% to $2.078 billion but posted a $740 million net loss.

One Prints Cash. The Other Prints Debt.

Meta is a self-funding machine. Advertising revenue climbed 33% to $55.02 billion, ad impressions rose 19%, and pricing per ad jumped 12%. That funded $19 billion in Q1 capex without touching the balance sheet. Mark Zuckerberg framed the buildout around “personal superintelligence” and the debut model from Meta Superintelligence Labs.

CoreWeave’s story is different. CEO Michael Intrator called it “the strongest bookings quarter in CoreWeave’s history,” with backlog approaching $99.4 billion and active power crossing 1 GW. Yet capex hit $7.695 billion, roughly 370% of quarterly revenue, and interest expense doubled to $536 million. Total liabilities now sit at $50.81 billion against $55.57 billion in assets.

The $21 Billion Handshake Hides a Structural Problem

Meta signed a $21 billion multi-year compute partnership with CoreWeave through 2032, which explains why CoreWeave’s backlog looks so fat. The catch is what Bloomberg reported about “Meta Compute,” an initiative to turn Meta’s in-house AI infrastructure into a public cloud that rents out excess bare-metal GPU capacity. Meta’s largest customer relationship with CoreWeave could morph into direct competition inside the same contract window.

That matters because CoreWeave’s pricing power rests on GPU scarcity. If Meta redirects even a slice of its $145 billion capital expenditure cycle into rentable capacity, the neocloud scarcity premium erodes. Insider selling ahead of the earnings report and the securities class action alleging concealed data center delays aren’t helping.

What I’m Watching Into Q2

Meta reports Q2 on July 29, and I want to see whether ad pricing holds while capex accelerates. For CoreWeave, the question is margin trajectory. Operating margin has flipped from +3.8% in Q3 2025 to -6.9% in Q1 2026 even as revenue exploded. That’s the wrong direction.

Why I’d Rather Own Meta Here

For me, Meta is the cleaner bet. It trades at a P/E near 21 with 57 buy ratings and a consensus target of $828.17 against a $600.29 quote. CoreWeave analysts still see upside to $142.29, but the setup only works if AI GPU pricing stays tight. If you believe Meta Compute lands as advertised, that assumption cracks. On the data, Meta’s self-funded balance sheet looks better positioned than CoreWeave’s debt-financed buildout if GPU pricing softens.

Contact [email protected] for any questions or corrections.