.png){kind=logo}
Meta Platforms stock moved higher on Friday as investors reacted to new details on artificial intelligence revenue plans, data center spending and model pricing. Shares rose about 6% during the session and gained nearly 15% for the week, marking the stock’s best weekly performance since early 2024.
The rally erased Meta’s losses for the year. The stock had faced pressure from concerns over rising capital spending and slower returns from AI investment. The latest updates gave Wall Street more data on how Meta may turn heavy infrastructure spending into new sales outside advertising.
Meta Chief Executive Mark Zuckerberg said the company is exploring the option of renting AI computing power to outside customers. He said demand for Meta’s compute capacity is ‘so high that it may make sense’ to sell part of that access through a cloud business.
The plan is still at an early stage. Meta has not given a final structure for the service. It may offer access to AI chips and servers, or it may host models for companies that need more computing power. This would place Meta in a market led by Amazon Web Services, Microsoft Azure and Google Cloud.
AI companies have said demand for computing capacity is rising as users adopt chatbots, coding tools and agentic systems. Meta already spends heavily on data centers for Facebook, Instagram, ads and AI tools. A cloud offering may help the company earn revenue from capacity that supports the same build-out.
Meanwhile, Meta released Muse Spark 1.1, a new AI model built for coding and agentic workloads. The company also introduced pricing for developer access. Meta plans to charge $1.25 per million input tokens and $4.25 per million output tokens.
This price is below Anthropic’s Claude Opus 4.8, which costs $5 per million input tokens and $25 per million output tokens. Meta’s pricing may appeal to developers looking for cheaper tools for coding, workflow automation and AI agents.
Tokens measure pieces of text used by AI models. Input tokens cover prompts and commands from users. Output tokens cover the response generated by the model. Lower token prices can help developers manage costs when apps process large volumes of text.
Meta also released Muse Image earlier in the week. The image model targets creators, advertisers and subscription products. Together, Muse Image and Muse Spark 1.1 show Meta is building AI tools for both business users and developers.
Moreover, Meta announced a new data center in Canada as part of its global infrastructure expansion. The project will be located in Alberta and will add to Meta’s large data center network. It also supports the company’s growing need for AI training and model delivery.
Meta is also moving forward with Iris, its custom AI chip. Reports say production may start in September. The chip forms part of Meta’s effort to lower costs and gain more control over hardware used across its AI systems.
Capital spending remains a key issue for investors. Meta previously guided 2026 capital expenditure as high as $145 billion. That figure raised questions over how fast AI spending would produce revenue.
Bank of America analyst Justin Post wrote that Meta ‘may have engineered significant cost savings’ in data center capacity. The comment helped support the view that Meta may manage its AI build-out at a lower cost than expected.
Also Read: Tencent Leads $2 Billion Manus Buyback Talks After China Blocks Meta Deal
