The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Meta is reported to have signed a six-year Google Cloud contract worth more than $10 billion, adding another source of computing infrastructure as it expands for AI. The agreement’s reported services include servers, storage and networking, but the public reporting does not say which workloads or regions will use them.
What the Google Cloud deal reportedly includes
Reuters reported on August 21, 2025, that Google had struck a six-year cloud-computing agreement with Meta valued at more than $10 billion. Reuters said Meta would use Google Cloud servers, storage, networking and other services. The Information separately reported the six-year term and value, describing the contract as one of the largest known in Google Cloud’s history.
These details come from reporting that cited people familiar with confidential discussions. No public contract, joint announcement, pricing schedule or workload allocation was identified in the available reporting. The value should therefore be treated as a reported contract amount, not as a published breakdown of what Meta will buy each year or in each location.
How the agreement could add AI capacity
Cloud infrastructure gives Meta access to computing resources and related services supplied by Google, rather than requiring every increment of capacity to come from Meta-owned data centers. The named services—servers, storage and networking—are the basic infrastructure used to run workloads and move or retain data. The reporting does not specify how much capacity Google will provide, which AI systems will use it, or whether Google will supply particular AI accelerators.
#1 Best Overall
The agreement is one part of a broader expansion, not a standalone measure of Meta’s AI capacity. Reuters later reported, citing an internal Meta memo, that the company planned to double its total computing capacity again to reach 14 gigawatts in 2027 and expected to spend as much as $145 billion on AI infrastructure in 2026. Those are company-wide plans, not figures for the Google contract; the reports do not identify the contract’s share of either total.
Why Meta is using multiple cloud providers
Meta has historically operated its own data centers, while also signing cloud agreements with Amazon Web Services, Microsoft Azure, Oracle and CoreWeave, according to The Information. Adding Google Cloud broadens that supplier mix while Meta continues its own infrastructure build-out.
A wider set of providers can give a company additional capacity and more options when sourcing services. The Information suggested the Google agreement could also give Meta more leverage in negotiations with AWS and Azure and help it serve users around the world. That is an interpretation of the deal’s possible business value; public reporting does not disclose comparative prices, performance, or a regional deployment plan.
What the deal means for Google Cloud
For Google Cloud, the reported agreement would bring in a major technology-company customer and add to its large AI-related contracts. The Information characterized it as one of the largest known agreements in Google Cloud’s 17-year history. The reported terms do not reveal which Google Cloud products or infrastructure types will account for most of the contract.
Recommended Free Tools
For Meta, the deal adds an external supplier; for Google, it represents a significant reported customer commitment. Without published workload, capacity or pricing details, the agreement cannot establish that one provider will replace another or that any particular cloud service is more capable or cost-effective.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Meta’s AI build-out also depends on physical infrastructure
Cloud services are only one piece of a large computing expansion. Reuters reported that Meta’s long-term supply agreements also included Samsung Electronics for memory chips, SanDisk for flash storage and Sumitomo Electric for fiber-optic equipment. These agreements illustrate the range of components needed to expand computing systems, but the reporting does not quantify how much capacity each supplier will support.
Quick Recap
Best Value
What remains unknown
- Workload allocation: The reports do not say which Meta products, AI models or internal tasks will run on Google Cloud.
- Infrastructure mix: Servers, storage and networking are named, but the reports do not specify the role of AI accelerators or other hardware.
- Location and schedule: No regional deployment plan or detailed capacity timeline for this contract has been made public in the cited reporting.
- Commercial terms: The reported total value and term are known, but annual payments, pricing structure and service-level commitments are not.
- Share of Meta’s capacity: No reported estimate establishes what portion of Meta’s total computing workload or infrastructure will come from Google.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
