Anthropic Commits $11.6 Billion to Akamai Cloud Deal
Seven-year deal targets general-purpose CPU compute as AI agent workloads scale.
Anthropic has agreed to spend $11.6 billion over seven years on Akamai Technologies' cloud infrastructure, dramatically expanding a previous $1.8 billion commitment. The agreement highlights growing compute requirements beyond GPU clusters as autonomous AI agent workflows scale. To meet the contract's capacity demands, Akamai plans to spend $5.5 billion building out infrastructure and an additional $1.7 billion on advance component procurement.
Key details
Under the seven-year contract, Anthropic's financial commitment is tied to Akamai meeting specific delivery and service-availability milestones. Akamai expects the contract to generate between $150 million and $300 million in revenue during the second half of 2027, scaling to an annual run rate of approximately $1.7 billion by late 2028.
To support the massive compute capacity, Akamai plans to invest $5.5 billion in capital expenditure, while allocating $1.7 billion in 2026 capital spending for advance hardware components, including memory modules. A key focus of the deployment is general-purpose CPU capacity, which handles code execution, browser interaction, and orchestration tasks required by scaling AI agents.
As part of the structure, Akamai issued Anthropic warrants for preferred stock convertible into up to 7.7 million common shares (approximately 5% of outstanding stock) at $111.33 per share. An initial 2% vests upon Anthropic's first payment, with the remaining 3% unlocked in $3 billion spending increments, allowing the total contract value to potentially reach $20 billion.
Why this matters
The multi-billion-dollar contract underlines how the hardware economics of AI are expanding beyond GPU accelerators to encompass massive general-purpose CPU infrastructure. As AI applications transition from simple prompt-response interactions to autonomous agentic workflows that continuously execute code and handle environment interactions, CPU and memory demands are scaling exponentially, adding billions to infrastructure bills.
Context
This deal represents Akamai's largest contract in company history and reflects a broader trend of AI labs locking in multi-gigawatt and multi-billion-dollar compute leases across diverse cloud providers. While major hyperscalers like AWS and Google Cloud remain primary training partners for Anthropic, specialized cloud and edge providers like Akamai are securing critical roles in handling auxiliary compute, inference, and agent orchestration workloads.
Risks and open questions
A central question is whether the massive capital investments required by cloud providers—such as Akamai's $5.5 billion buildout and $1.7 billion upfront component spend—will yield sustainable margins if AI application demand shifts. Furthermore, the performance-contingent nature of the contract means any infrastructure delivery delays could alter projected revenue timelines.
What happens next
Akamai will begin expanding its global cloud infrastructure footprint and procuring component inventory through late 2026, with initial revenue recognition scheduled for the second half of 2027. Investors and industry analysts will track whether Anthropic exercises its warrants to expand the contract toward its $20 billion maximum cap.
Source: TechCrunch Published on AI Usage Global, author: AUG Bot
