Meta reportedly plans to launch cloud infrastructure business

On Wednesday, Bloomberg reported that Meta is developing plans for a cloud infrastructure business to sell access to AI compute power—defined as computational power used for AI workloads—and models. The initiative, reportedly dubbed Meta Compute, would mark a significant shift for the social media giant. By offering these services, Meta would enter direct competition with major cloud infrastructure providers, including Amazon Web Services and Microsoft Azure. The new business line is reportedly being led by Meta’s head of infrastructure Santosh Janardhan, Meta Superintelligence Labs leader Daniel Gross, and company president Dina Powell McCormick.

The scale of Meta’s infrastructure buildout is massive. The company has committed to spending $182.9 billion on AI infrastructure, which underpins what some industry observers characterize as trillion-dollar bets on artificial intelligence. This investment includes large-scale data center projects in Louisiana and Ohio. The Ohio project, which CEO Mark Zuckerberg has described as being the size of Manhattan, is expected to come online this year. Zuckerberg previously indicated in May that a cloud computing business is “definitely on the table” as a pathway to generate returns on these massive capital expenditures. Currently, Meta does not break out revenue from Meta AI or its Llama models in its earnings, and executives have mostly emphasized internal corporate uses of AI in public statements.

To structure this new business, Meta is reportedly considering models similar to those of specialized cloud provider CoreWeave and aerospace firm SpaceX. In early May, SpaceX, acting through xAI, signed a deal with Anthropic to lease compute capacity at its Colossus 1 data center, and has since signed similar leases with Google and Reflection AI. Under its planned model, Meta is reportedly considering selling access to raw computational capacity as well as hosting various AI models on its infrastructure. This would include Muse Spark, which is a closed-weight model—meaning its weights are proprietary—in contrast to Meta’s Llama family of open-weight models, where the underlying weights are made publicly available.

Why it matters

Meta is looking to monetize its massive AI infrastructure investments by launching a cloud business, potentially competing with major cloud providers like AWS and Azure.