Meta sees 2026 capex surging up to $135 billion on AI infrastructure push

Meta sees 2026 capex surging up to $135 billion on AI infrastructure push

The Facebook and Instagram owner reported revenue of $59.89 billion for the fourth quarter and $200.97 billion for full-year 2025, representing year-over-year growth of 24% and 22%, respectively.

Advertisement
MetaMeta
Business Today Desk
  • Jan 29, 2026,
  • Updated Jan 29, 2026 11:40 AM IST

Meta Platforms Inc. said it expects capital expenditure to climb to between $115 billion and $135 billion in 2026, a sharp increase from the $72.2 billion spent in 2025, as the company accelerates investment in data centres and specialised hardware to support advanced artificial intelligence (AI).

Chief Executive Officer Mark Zuckerberg told analysts on a conference call that 2026 would be a “big year” for Meta’s AI ambitions, framing “personal superintelligence”, machines that could eventually surpass human cognitive abilities, as central to the company’s long-term strategy.

Advertisement

The Facebook and Instagram owner reported revenue of $59.89 billion for the fourth quarter and $200.97 billion for full-year 2025, representing year-over-year growth of 24% and 22%, respectively.

Even as Meta pours resources into AI, it is trimming parts of its legacy bets. The company confirmed it is cutting about 10% of staff in its Reality Labs division, which oversees metaverse initiatives, as it shifts focus toward AI-powered wearables.

Chief Financial Officer Susan Li said Meta expects first-quarter 2026 revenue in the range of $53.5 billion to $56.5 billion.

“We expect first quarter 2026 total revenue to be in the range of $53.5–56.5 billion. Our guidance assumes foreign currency is an approximately 4% tailwind to year-over-year total revenue growth, based on current exchange rates,” Li said.

Advertisement

Meta forecast total expenses for full-year 2026 of between $162 billion and $169 billion.

Headcount stood at 78,865 as of December 31, 2025, up 6% from a year earlier, while long-term debt totalled $58.74 billion.

Total costs and expenses rose 40% year over year in the fourth quarter to $35.15 billion, and climbed 24% for full-year 2025 to $117.69 billion, reflecting the growing cost of scaling AI infrastructure.

Meta Platforms Inc. said it expects capital expenditure to climb to between $115 billion and $135 billion in 2026, a sharp increase from the $72.2 billion spent in 2025, as the company accelerates investment in data centres and specialised hardware to support advanced artificial intelligence (AI).

Chief Executive Officer Mark Zuckerberg told analysts on a conference call that 2026 would be a “big year” for Meta’s AI ambitions, framing “personal superintelligence”, machines that could eventually surpass human cognitive abilities, as central to the company’s long-term strategy.

Advertisement

The Facebook and Instagram owner reported revenue of $59.89 billion for the fourth quarter and $200.97 billion for full-year 2025, representing year-over-year growth of 24% and 22%, respectively.

Even as Meta pours resources into AI, it is trimming parts of its legacy bets. The company confirmed it is cutting about 10% of staff in its Reality Labs division, which oversees metaverse initiatives, as it shifts focus toward AI-powered wearables.

Chief Financial Officer Susan Li said Meta expects first-quarter 2026 revenue in the range of $53.5 billion to $56.5 billion.

“We expect first quarter 2026 total revenue to be in the range of $53.5–56.5 billion. Our guidance assumes foreign currency is an approximately 4% tailwind to year-over-year total revenue growth, based on current exchange rates,” Li said.

Advertisement

Meta forecast total expenses for full-year 2026 of between $162 billion and $169 billion.

Headcount stood at 78,865 as of December 31, 2025, up 6% from a year earlier, while long-term debt totalled $58.74 billion.

Total costs and expenses rose 40% year over year in the fourth quarter to $35.15 billion, and climbed 24% for full-year 2025 to $117.69 billion, reflecting the growing cost of scaling AI infrastructure.

Read more!
Advertisement