Meta expects its 2026 capex to be $135B-$145B, up from $130B-$145B previously, and forecasts Q4 revenue of $73B-$76B, vs. $73.6B est.; META jumps 3%+ pre-market
Riley Griffin /Bloomberg:
Context & Ripple Effects
Meta began 2026 with a $115B-$135B infrastructure budget tied to Superintelligence Labs, then lifted the range to $125B-$145B in April in an earlier capex forecast increase. The latest revision raises the floor again while preserving the ceiling.
The spending path aligns with Meta's disclosed Prometheus supercluster and Muse model work. Its confirmed plan for nonvoting Class C shares, effective January 2027, also puts a dedicated financing mechanism alongside the expanding AI budget.
First-order effects
- Meta is committing to at least $135B of 2026 capital expenditure, increasing the resources available for AI infrastructure even as its Q4 revenue midpoint sits below the cited analyst estimate.
- The higher spending floor strengthens the financial rationale for Meta's planned Class C stock issuance to fund AI model development.
Second-order effects
- Meta's capital plan concentrates more of its AI strategy in large-scale compute assets such as Prometheus, raising the importance of converting that capacity into model progress and product value.
- A funding structure that preserves voting control while raising capital gives Meta a way to sustain unusually large AI outlays without relying solely on operating cash flow.
Third-order effects
- If other public AI developers pair escalating infrastructure budgets with bespoke financing, AI competition will increasingly turn on access to capital structures as well as research and computing capacity.
The trend: The story is part of an AI infrastructure capital cycle in which public platforms raise spending commitments and adapt their financing to support frontier-model compute.