Meta is developing plans for a cloud infrastructure business that would sell access to AI compute power and models, according to a Bloomberg report published Wednesday.

The move would put Meta in direct competition with established cloud providers Amazon Web Services, Google Cloud, and Microsoft Azure. Meta has committed to spending $182.9 billion on AI infrastructure in coming years as of the first quarter.

Meta's strategy mirrors SpaceX's recent approach through xAI. In May, SpaceX signed a deal with Anthropic to buy out all compute capacity at SpaceX's Colossus 1 data center, followed by similar agreements with Google and Reflection AI.

The timing suggests that owning data centers, rather than building the best AI models, may determine the winners in the AI race. Meta's massive infrastructure investments include ongoing projects in Louisiana and Ohio, with the Ohio facility expected to be the size of Manhattan when it comes online this year.

Unlike Google and OpenAI, Meta hasn't seen significant external demand for its AI models and services. The company doesn't break out revenue from Meta AI or its Llama open-weight model family in earnings reports.

Meta executives have primarily emphasized internal corporate uses of AI in public statements, suggesting the company's AI efforts don't yet represent a material standalone revenue stream.

Two potential revenue models

Meta may follow CoreWeave's approach by selling access to raw compute capacity, Bloomberg reported. The company is also considering an AWS-style model that would host various AI models on its infrastructure.

This could include Meta's recently launched closed-weight model, Muse Spark, alongside third-party offerings.

Some analysts have questioned whether the rapid buildout of AI infrastructure creates a bubble dependent on quickly depreciating chips. Others doubt AI companies can generate sufficient end-user revenue to justify trillion-dollar infrastructure investments.

Meta's cloud business plans come as the company seeks returns on its colossal AI spending while demand for its own models remains limited compared to competitors.