Meta Platforms is building a cloud computing business. The company that spent the last few years pouring money into AI infrastructure apparently looked at all that computing power and thought, “Why not rent it out?”

The initiative, called Meta Compute, would let third-party customers tap into Meta’s excess AI computing capacity and hosted AI models. Meta shares jumped roughly 8-9% on the news.

What Meta Compute actually looks like

Meta Compute isn’t just about selling raw server time. The offering is expected to include both computing power and access to advanced AI models, including Muse Spark, a service compared to Amazon’s Bedrock platform.

The effort is being led by a trio of executives: Santosh Janardhan, who heads Meta’s infrastructure; Daniel Gross from Meta Superintelligence Labs; and president Dina Powell McCormick.

Meta’s 2026 capital expenditure guidance sits between $125 billion and $145 billion. CEO Mark Zuckerberg tipped his hand back in May 2026, suggesting a cloud business was a feasible path for recouping investments in AI superintelligence. The July announcement formalized what was already being discussed internally.

Picking a fight with the biggest names in tech

Meta Compute puts the company on a direct collision course with Amazon Web Services, Microsoft Azure, and Google Cloud.

The comparison to AWS Bedrock is telling. Amazon’s managed AI service lets developers access foundation models from multiple providers without managing the underlying infrastructure. If Muse Spark follows a similar playbook, Meta would be offering something that directly undercuts one of AWS’s fastest-growing product lines.

What this means for investors

The 8-9% stock jump tells you everything about what investors were worried about. Meta’s AI spending has been the elephant in the room for quarters, with analysts questioning whether the company could ever generate returns proportional to its infrastructure bets. Meta Compute is the first concrete answer to that question.

If Meta is spending up to $145 billion on AI infrastructure in a single year and can monetize even a fraction of the excess capacity through cloud services, the revenue opportunity is substantial.

The risk is execution. Meta has a history of pivoting into new business lines with mixed results. The metaverse push, rebranding the entire company around a concept that hasn’t materialized at scale, is the obvious cautionary tale. Cloud computing requires a fundamentally different organizational muscle: enterprise sales, service-level agreements, uptime guarantees, compliance certifications, and the kind of customer support that a consumer-focused company has never needed to build.