Meta plans AI compute cloud service to compete with AWS, Google, Microsoft
According to Bloomberg, Meta is developing plans for a cloud infrastructure business to sell access to its AI compute power and models.
Meta is reportedly preparing to enter the cloud infrastructure market by selling access to its AI compute resources and proprietary models, according to a Bloomberg report cited by TechCrunch. The proposed business line, referred to as Meta Compute, places the company in direct competition with established cloud providers such as Amazon Web Services, Google Cloud, and Microsoft Azure.
The initiative is led by infrastructure head Santosh Janardhan, Meta Superintelligence Labs leader Daniel Gross, and president Dina Powell McCormick, according to TechCrunch. This follows statements made by CEO Mark Zuckerberg in May indicating that a cloud computing business was definitely on the table as a method to secure returns on massive investments directed toward developing AI superintelligence.
According to TechCrunch, Meta's strategy may involve copying CoreWeave's business model to sell access to raw compute capacity, alongside following AWS by selling access to various AI models hosted on its infrastructure, including its recently launched closed-weight model, Muse Spark. Bloomberg reported these details regarding the potential structure of the new enterprise.
Financial commitments to the underlying infrastructure remain substantial. TechCrunch noted that by the conclusion of the first quarter, the company had lined up $182.9 billion in expenditures directed at future AI infrastructure, which includes massive ongoing projects located in Louisiana and Ohio. The Ohio project, described by Zuckerberg as being the size of Manhattan, is expected to come online this year.
Despite these heavy investments, TechCrunch reported that Meta has not seen significant demand for its own AI models and services when compared to competitors like Google and OpenAI. Executives have largely emphasized internal corporate uses for AI in public statements, and the company does not break out revenue from Meta AI or its Llama open-weight model family in its earnings reports.
The push to monetize excess compute follows similar moves across the industry. According to TechCrunch, SpaceX announced plans via xAI to sell off compute capacity at its Colossus 1 data center, signing leases with Anthropic, Google, and Reflection AI. Analysts and skeptics, however, have questioned whether the race to build AI infrastructure is creating a bubble relying heavily on rapidly depreciating chips and whether companies can generate sufficient end-user revenue to justify trillion-dollar bets.
Sources
- 01 Meta, like SpaceX, looks to turn excess AI compute into cash — TechCrunch