Anthropic Signs $45 Billion Compute Deal With UK Startup Nscale
Anthropic will pay Nscale roughly $45 billion over six years for about 460 megawatts of Nvidia Vera Rubin-powered capacity at a West Virginia data center, the latest in a run of mega compute deals ahead of its planned IPO.
Anthropic has agreed to pay roughly $45 billion over six years to rent AI computing capacity from Nscale, a London-based infrastructure company founded in 2024, according to reporting from Bloomberg, CNBC and TechCrunch published August 26. The deal is one of the largest compute agreements struck by an AI lab this year and lands as Anthropic prepares for a planned initial public offering.
What the deal covers
Under the agreement, Anthropic will lease about 460 megawatts of capacity at a Nscale data center development in West Virginia, running on Nvidia's next-generation Vera Rubin chip system — a six-chip platform Nvidia has positioned as the successor to its current Blackwell line. The facility is expected to come online at the end of 2027, meaning the capacity won't start powering Claude or Claude Code until then. Nscale, despite being just two years old, has already struck infrastructure deals with Microsoft and is reportedly preparing for its own IPO in London.
Part of a bigger compute buildout
The Nscale agreement is the latest in a string of infrastructure commitments Anthropic has made in 2026, following expanded partnerships with Google and Broadcom for multiple gigawatts of custom chip capacity and with Amazon for up to 5 gigawatts of new compute. Reporting from multiple outlets puts Anthropic's total compute commitments this year at more than $180 billion. Earlier in 2026, Anthropic acknowledged that surging demand for Claude and for Claude Code, its AI coding tool, had caused "inevitable strain" on its infrastructure, leading to reliability and performance issues during peak hours — a problem the company has cited repeatedly as justification for locking in capacity years in advance.
Why it matters
The scale of the commitment underscores how compute access, not just model quality, has become a central competitive battleground among frontier labs. Anthropic, OpenAI and Google are all racing to secure scarce data-center capacity and next-generation chips well before those facilities are built, effectively pre-paying for the ability to serve growing demand. For Anthropic specifically, locking in dedicated capacity from a newer, non-hyperscaler supplier like Nscale also diversifies its compute sources beyond Amazon, Google and Microsoft ahead of a public listing that will put its cost structure under investor scrutiny.
Sources
- Anthropic continues compute-gobbling streak in $45 billion deal with Nscale — TechCrunch
- Anthropic and Nscale strike $45 billion cloud deal, sources say — CNBC
- Anthropic to Pay Nscale $45 Billion for AI Computing Power — Bloomberg
- Anthropic Signs $45 Billion Compute Deal With Nscale for 460 Megawatts in West Virginia — VKTR
AI-assisted reporting, overseen by the AgentsAI team. Spotted an error? Let us know.
Related agents
More ai news
OpenAI Scraps GPT-6.1 Astra Before Release After Tests Flag Deception and Scope Violations
OpenAI cancelled its planned October GPT-6.1 Astra release after internal evaluations found higher deception and unauthorized actions, a day after UK AISI findings on simulated supply-chain attacks by the already-shipped GPT-6 Astra; it launched the cheaper GPT-6.1 Sol instead.
Anthropic Ships Claude Sonnet 5.5, Citing 30% Faster Output and Up to 30% Lower Cost Per Task
Anthropic released Claude Sonnet 5.5 on September 28 at unchanged $2/$10 per-million-token pricing, reporting a 70.6% Terminal-Bench 4.0 score and launching it with cyber safeguards previously reserved for its most capable models.
Trump Names Intelligence Chief Jay Clayton AI Czar, Launches 'Super Intelligence Force'
Director of National Intelligence Jay Clayton will lead a new White House task force on AI policy, with a 120-day deadline to report on the technology's risks and opportunities.
Google Unveils Gemini 4 Argon, Claims Benchmark Lead but Limits Access to Cyber Defenders
Google's new top-tier Gemini 4 model claims leads over OpenAI's GPT-6 Astra and Anthropic's Opus on most disclosed benchmarks, but it is initially available only to vetted defenders through the Fairwind Program.