# Akamai locks in $11.6B as Anthropic bets on CPUs _Akamai signed its largest deal ever, an $11.6B seven-year CPU cloud commitment with Anthropic that could reach $20B._ **Published:** 2026-09-26 **Source:** https://www.startuphub.ai/ai-news/public-companies/2026/akamai-locks-in-11-6b-as-anthropic-bets-on-cpus --- Akamai announced an $11.6 billion, seven-year commitment from [Anthropic](https://www.akamai.com/newsroom/press-release/akamai-announces-11-6-billion-multi-year-agreement-with-anthropic-to-support-growing-demand) to power cloud infrastructure for rising demand around its models, according to [the announcement](https://www.akamai.com/newsroom/press-release/akamai-announces-11-6-billion-multi-year-agreement-with-anthropic-to-support-growing-demand). This deal transforms Akamai from a pure delivery network into a main AI compute provider. The agreement leans on CPUs rather than GPUs and could climb toward $20 billion if Anthropic taps extra capacity, per [TechCrunch](https://techcrunch.com/2026/09/25/anthropic-to-pay-akamai-11-6-billion-over-seven-years-in-cloud-deal/), which also notes Akamai will award Anthropic warrants covering up to 5% of its stock as part of an unusual equity kickback linked to spending. This structure reshapes how teams building large-scale Claude models allocate resources. Workloads that don't need accelerators can move to Akamai Cloud to chase lower prices and better availability, while GPU clusters remain dedicated to training and cutting-edge inference. Operators will need to split deployments, rework autoscaling and routing rules to treat Akamai as a core region instead of an edge cache. Since acquiring Linode and expanding distributed compute, Akamai has pushed its cloud platform hard, yet it still battles AWS, Google Cloud and Azure for enterprise contracts and faces competitors like [CoreWeave](https://www.startuphub.ai/ai-news/artificial-intelligence/2026/coreweave-6-billion-lancaster-data-center) and Lambda for AI-native capacity. A flagship customer paying around $1.65 billion annually provides steady utilization, justifying data center builds and giving Akamai a foothold convincing buyers who've already tried its delivery and security services. The published release omits details on pricing, prepayment or early termination, and Akamai's own press page didn't go beyond the headline figure at launch. If the CPU focus proves right, cost-conscious inference operations will thrive. Otherwise, Akamai keeps its seven-year commitment intact. --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.