Meta has signed a six-year, $10 billion agreement with Google Cloud to expand its artificial intelligence infrastructure, according to reports from Bloomberg and Reuters. The move highlights Meta’s growing need for scalable compute and storage as it accelerates AI development across Facebook, Instagram, and WhatsApp.
Under the deal, Meta will rely on Google Cloud’s servers, storage, and networking services, supplementing its own large-scale data center projects like the upcoming Hyperion campus in Louisiana.
Why It Matters
- AI Growth: Training large language models such as Meta’s Llama requires vast GPU resources, straining even hyperscaler-owned facilities.
- Cloud Competition: Meta has historically leaned on Amazon Web Services and Microsoft Azure. This deal positions Google Cloud as a new critical partner.
- Industry Context: Google Cloud has also landed OpenAI as a customer, strengthening its positioning as an AI-first cloud provider.
Balancing Build and Buy
While Meta continues to invest in its own AI-optimized data centers, analysts note that the rapid scale of AI demand has forced even hyperscalers to seek external cloud capacity.
Meta’s latest $1 billion Kansas City Data Center, which went live in August 2025, reflects this dual approach — building massive in-house facilities while also securing external partnerships. Read more here.
“Meta is one of the most sophisticated data center operators in the world, but the speed of AI deployment today means that even hyperscalers are looking for creative solutions,” said Chuck Marvin of Thunderhead Energy Solutions in comments to CNET.
The Bigger Picture
Meta’s capital expenditures for 2025 are projected between $66–72 billion, with a significant portion directed at AI and data centers. The Google Cloud deal provides additional short-term infrastructure while Meta’s in-house projects catch up.
For Google Cloud, the partnership strengthens its push to challenge AWS and Azure, positioning itself as a go-to provider for enterprise AI workloads.