Meta is developing plans for a cloud infrastructure business that sells access to AI compute power and models, according to a Bloomberg report on 1 July 2026 [1][2]. The move would put Meta in direct competition with the big cloud providers Amazon Web Services, Google Cloud, and Microsoft Azure [1].

An AI cloud business sells access to GPU compute capacity and models hosted in data centers the seller owns, so customers pay to run AI workloads without building their own infrastructure [1].

Why now

Meta's step comes weeks after SpaceX, through xAI, announced similar plans. In early May 2026, SpaceX signed a deal with Anthropic to buy out all the compute capacity at its Colossus 1 data center, and later signed similar leases with Google and Reflection AI [1]. Meta following the same pattern is a signal that the AI race may be decided by who owns the data centers [1].

How big Meta's bet is

As of the end of the first quarter of 2026, Meta had committed 182.9 billion US dollars to AI infrastructure in the coming years, including massive ongoing projects in Louisiana and Ohio [1][3]. The Ohio project, which Zuckerberg said would be the size of Manhattan, is expected to come online this year [1].

Meta has not seen significant demand for its own AI models and services. The company does not break out revenue from Meta AI or from Llama, its open-weight model family, in its earnings, and executives have mostly emphasized internal corporate uses of AI in public statements [1]. That means Meta's AI endeavors do not yet represent a material standalone revenue line [1].

The business model: CoreWeave and AWS

To get a return on its massive spend, Meta could copy CoreWeave's model and sell access to raw compute capacity, according to Bloomberg [1][2]. Meta is also considering following AWS's lead by selling access to various AI models, including its recently launched closed-weight model Muse Spark, hosted on its infrastructure [1].

The new business line will be part of an initiative dubbed Meta Compute, led by head of infrastructure Santosh Janardhan, Meta Superintelligence Labs leader Daniel Gross, and president Dina Powell McCormick [1]. The move confirms Zuckerberg's May 2026 statements that a Meta cloud computing business is firmly on the table [1][4].

The contested risks

Some skeptics warn that the race to build AI infrastructure is creating a bubble that rests on chips whose value depreciates quickly [1]. Others question whether AI companies can generate enough end-user revenue to justify trillion-dollar bets [1]. Those concerns have not stopped Meta from investing heavily in AI compute infrastructure [1].

What it means for AI builders in Asia

For AI builders in Indonesia and Asia, three signals are worth noting. New cloud AI competition could lower compute prices as excess capacity is released to the market [1]. Alternative infrastructure beyond AWS, Google Cloud, and Azure opens procurement options for teams running agents all day [1]. Data center ownership is now a source of advantage, so the players with the largest GPU reserves hold stronger bargaining power [1].

Frequently asked questions

What is the Meta Compute cloud business?

Meta's planned cloud infrastructure business to sell access to AI compute power and models hosted in its own data centers, reported by Bloomberg on 1 July 2026. The initiative is called Meta Compute and is led by Santosh Janardhan, Daniel Gross, and Dina Powell McCormick [1][2].

How big is Meta's AI infrastructure commitment?

As of the end of the first quarter of 2026, Meta has committed 182.9 billion US dollars to AI infrastructure in the coming years, including major projects in Louisiana and Ohio. The Ohio project is expected to come online this year [1][3].

How does Meta compare with SpaceX and xAI?

In early May 2026, SpaceX through xAI signed a deal with Anthropic to buy out all compute capacity at the Colossus 1 data center, then similar leases with Google and Reflection AI. Meta is following the same pattern, selling excess compute capacity to the market [1].

What business model is Meta copying?

Meta could copy CoreWeave by selling access to raw compute capacity, and follow AWS by selling access to various AI models including Muse Spark hosted on its infrastructure [1].

Sources

  • 1. TechCrunch, Rebecca Bellan, Meta, like SpaceX, looks to turn excess AI compute into cash. https://techcrunch.com/2026/07/01/meta-like-spacex-looks-to-turn-excess-ai-compute-into-cash/
  • 2. Bloomberg, Meta is building a cloud business to sell excess AI compute. https://www.bloomberg.com/news/articles/2026-07-01/meta-is-building-a-cloud-business-to-sell-excess-ai-compute
  • 3. SEC filing, Meta Platforms Q1 2026. https://www.sec.gov/Archives/edgar/data/1326801/000162828026028526/meta-20260331.htm
  • 4. CNBC, Zuckerberg says Meta starting cloud business on the table. https://www.cnbc.com/2026/05/27/mark-zuckerberg-says-meta-starting-cloud-business-on-the-table.html