Meta reported revenue of $60.80 billion for the second quarter of 2026, up 28% from $47.52 billion the same period last year. That beat what Wall Street analysts expected. But earnings per share, a measure of profit for each share of stock, came in at $6.18, well below the $7.22 analysts predicted. The company shared these results on July 29, 2026. CNBC Meta Investor Relations

Mark Zuckerberg used the earnings call to lay out a much bigger plan for selling AI to other businesses. He told investors that Meta sees "a large enterprise opportunity to sell to businesses, including APIs, business agents, potentially selling compute directly, and other services that we're building for large customers." An API is a way for different software programs to communicate with each other. Selling API access means letting outside companies build their own products using Meta's AI technology. TechCrunch

The first step is to target advertisers. Zuckerberg said Meta will start by offering AI agents, software programs that can handle tasks on their own, that work across messaging apps. These agents would let businesses interact with customers through an AI interface. The payment model mirrors the ad business: "just like the ad system, effectively, we will get paid when we deliver results for those businesses." He described the push as "an extension of the sales and the partnerships that we have with many millions of advertisers and hundreds of millions of small businesses that use our platforms." TechCrunch

Beyond agents, Zuckerberg said Meta is building coding, development, and internal productivity tools and sees a "large opportunity to serve" external customers with them, both small and larger businesses. He acknowledged that selling to the enterprise was a "different muscle" than Meta has historically flexed. TechCrunch

Selling compute power was a recurring theme on the call. Compute refers to the processing power that runs AI systems, and Meta said it has the opportunity to sell it at "a significant premium over what we paid for it." In other words, Meta buys expensive computers and servers, then rents out their processing power to other companies at a markup. Zuckerberg cautioned that it "would be foolish" to "sell all of the compute and take a short-term profit," describing Meta's approach as a "portfolio" mixing long-term and short-term plans. Meta is projected to spend as much as $145 billion on AI infrastructure in 2026, part of a broader Big Tech collective spend projected to exceed $700 billion. TechCrunch Reuters

On the consumer side, Zuckerberg said that "as we get closer to personal superintelligence, we are going to need hardware that allows you to seamlessly interact with it." Meta plans to offer "personal AI agents" and AI smart glasses alongside its enterprise offerings. TechCrunch

TechCrunch also reported that Meta is using large language models, the same type of AI technology behind ChatGPT, to build out social apps including a Marketplace seller app, a Facebook Groups app called Forum, and a vibe-coded gaming app called Pocket. TechCrunch

The earnings call follows a quarter in which Meta unveiled an enterprise-focused AI business agent on June 3, 2026, aimed at automating daily business operations. Reuters In early July, Zuckerberg said AI agent development was progressing slower than expected. Reuters On the infrastructure front, Meta entered an agreement with Reliance Industries to lease an AI-enabled data center in Jamnagar, Gujarat, its first data center in India. Meta Data Centers Meta is also signing the EU AI Act Code of Practice on Transparency of AI-Generated Content. Meta Investor Relations

The earnings miss against a revenue beat is worth pausing on. Revenue growth at 28% year-over-year is strong by any measure, and Zuckerberg's framing in the press release that "AI is accelerating our core business today" is consistent with the top line. But the per-share shortfall signals that the cost of that acceleration, principally the massive spending on data centers and chips, is weighing on profitability faster than the new revenue streams can offset it. The $145 billion infrastructure projection for 2026 puts a fine point on this: Meta is spending at a scale that demands the enterprise monetization strategy Zuckerberg outlined to eventually generate returns well beyond advertising.

The enterprise ambition itself is notably broad. APIs, business agents, direct compute sales, coding and development tools, and internal productivity tools aimed at external customers cover a wide swath of the business software world. Meta is effectively outlining a move from a consumer-and-advertising company into an infrastructure-and-platform provider. The "different muscle" acknowledgment is candid; Meta has never operated a large enterprise sales organization, and the process of selling compute, APIs, and developer tools to large companies is fundamentally different from selling ad space to small businesses through a self-serve website.

The compute-selling idea is especially notable. If Meta sells compute at a significant premium over cost, it positions itself as an alternative to the big cloud providers, companies like Amazon Web Services, Google Cloud, and Microsoft Azure, for at least some workloads. The "portfolio" language, balancing short-term compute sales against long-term internal needs, suggests Meta intends to manage this carefully rather than flood the market. The risk is straightforward: selling compute profitably requires not just spare capacity but the reliability, compliance, and customer support that cloud customers expect.

On the consumer hardware side, the linkage between "personal superintelligence" and smart glasses is consistent with Meta's prior investment in wearable devices. The phrasing implies a timeline in which AI capabilities and hardware evolve together, with each enabling the other. Whether the market for AI smart glasses matures fast enough to justify the infrastructure investment is an open question, but Meta is clearly making the bet that people will interact with advanced AI through wearable devices rather than just screens.

The broader picture is a company with strong revenue momentum from its core advertising business, making a multi-pronged push to make money from AI across enterprise APIs, compute, developer tools, consumer agents, and hardware. The earnings miss shows that this transition is expensive. Zuckerberg's candor about both the opportunity and the execution challenges, from the "different muscle" of enterprise sales to the slower-than-expected pace of agent development, gives investors a clear-eyed picture of what lies between the current quarter and the returns Meta expects.