Meta's stock climbed more than 5% after the company announced plans to rent its AI computing power to third parties and launched a cheaper AI model with token pricing that undercuts competitors like Anthropic and OpenAI. The moves signal potential new revenue streams to offset Meta's heavy AI spending, which had weighed on the stock year to date.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Meta's stock rose more than 5% on Friday after CEO Mark Zuckerberg revealed the company is exploring renting its AI computing power to third parties. The company also announced its Muse Spark 1.1 AI model with pricing of $1.25 per million input tokens and $4.25 per million output tokens—substantially cheaper than Anthropic's Opus 4.8 model, which charges $5 per million input tokens and $25 per million output tokens.
Why it matters
Meta's stock had been down year to date due to concerns about rising capital expenditures and questions about returns on its massive AI investments. A new cloud computing revenue stream from renting data center capacity could offset those spending concerns. The aggressive token pricing also positions Meta to attract developers who want lower-cost AI model access, potentially expanding its foothold in the AI software market beyond its advertising business.
What to watch
Meta announced it will build a new data center in Canada—its 33rd such facility. The company has not disclosed how exactly it will sell its data center capacity, but it could either power AI models from competitors or sell direct access to its AI chips and servers similar to a cloud provider.
Meta's stock momentum reflects investor relief over a concrete plan to monetize its AI investments. The company has spent heavily on data center buildout—most recently announcing a new facility in Canada—but investors had grown anxious about when (or whether) that spending would generate returns. Two announcements on Thursday addressed that concern directly: a potential new revenue line from cloud services, and a pricing strategy designed to compete aggressively for developer adoption.
The token pricing is particularly significant because it undercuts not just Anthropic but positions Spark 1.1 as an economical choice for developers who don't require the highest-capability models. In AI markets, price competition typically follows capability maturation, suggesting Meta views Spark 1.1 as sufficiently capable for a broad slice of use cases. The data center rental pitch also taps into a real constraint: AI companies have told shareholders they are resource constrained and need more computing capacity. If Meta can capture some of that demand, it converts infrastructure spending into a revenue-generating service rather than a pure cost center.
No discussion yet for this article
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
