Meta✴ Platforms reported financial results for the first quarter of 2025, which ended March 31. The company’s key metrics beat analysts’ forecasts, and its Q2 outlook was in line with Wall Street expectations, sending the company’s shares up 5% in extended trading, CNBC reports.
Meta✴’s revenue increased 16% year-on-year to $42.31 billion, also beating the LSEG analyst consensus estimate of $41.40 billion. Net income jumped 35% to $16.64 billion from $12.37 billion a year earlier. Diluted earnings per share increased 37% to $6.43 from $4.37 last year, also beating the LSEG analyst consensus estimate of $5.28 per diluted share. Capital expenditures were $13.69 billion.
Meta✴ expects Q2 revenue to be between $42.5 billion and $45.5 billion, in line with analysts’ average forecast of $44.03 billion. The company also announced a planned reduction in total expenses in 2025 from the previous target of $114 billion to $119 billion to $113 billion to $118 billion.
Image source: Meta✴
At the same time, Meta✴ announced a decision to increase its capital expenditures for 2025 to $64 billion to $72 billion, up from its previous guidance of $60 billion to $65 billion. “This updated guidance reflects additional investments in data centers to support our AI efforts, as well as an increase in the expected cost of infrastructure equipment,” said CFO Susan Li. “The majority of our capital expenditures in 2025 will continue to be directed toward our core business,” she added.
When asked whether Meta✴’s higher capital expenditures on data center infrastructure are a result of new trade tariffs imposed by the U.S. administration, Susan Lee said, “The higher spending that we expect to see on infrastructure equipment this year is actually coming from suppliers that are sourcing from countries around the world. And there’s just a lot of uncertainty around that given the ongoing trade discussions.”
As The Register noted, it’s not just about tariffs. Li’s announcement also suggests that the company isn’t just paying more for AI data center components, but is also deploying more of them. Li said the revised capex forecast also reflects the company’s commitment to increasing the capacity of its AI data centers. The next-generation data centers are expected to support both internal training of base models and inference in the Meta✴ ecosystem, including smart glasses, LLM-powered assistants, and Family Sharing features.
She added that Meta✴ is working to improve the efficiency of its data center workloads. “Many of the innovations that have come out of our ranking work are aimed at improving the efficiency of our systems,” Li said. “This focus on efficiency helps us consistently deliver high returns on our core AI initiatives.” Meta✴ previously announced its intention to bring about 1 GW of compute capacity online in 2025 and have more than 1.3 million accelerators in place by the end of the year to train and serve AI models.