From $7.5B to $18B in 4 Months: Inside the AI Data Center Startup Everyone Wants a Piece Of

A 2.4x Jump in Four Months

$7.5 billion. That was Fluidstack's valuation in December 2025.

$18 billion. That's the number on the table four months later, in April 2026. Same company, same team, same business model -- but 2.4 times the price tag.

If you want to understand where the AI industry's center of gravity is shifting, this number tells you everything.

The Oxford Spinout That Moved to New York

Here's the deal: the biggest bottleneck in AI right now isn't models. It's the physical infrastructure to run them.

Since 2024, model sizes have exploded. GPT-5.4 runs trillions of parameters. Claude Mythos reportedly hit 10 trillion. Training and serving these models requires tens of thousands of high-end NVIDIA GPUs -- H100s, H200s -- but there simply aren't enough data centers to house them.

The Stanford AI Index 2026 report puts global AI data center power consumption at 29.6 gigawatts. That's enough to power the entire state of New York at peak demand. And it's still not enough.

Fluidstack saw this gap early. Originally spun out of Oxford University as a UK startup, the company made a decisive pivot in late 2025: it moved its headquarters from London to New York, pulled out of a large data center project in France, and went all-in on the US market where AI compute demand is most concentrated.

Metric	Value
Current negotiation valuation	$18B
Previous round valuation (Dec 2025)	$7.5B
Current round size	$1B
Lead investors	Jane Street, Situational Awareness
Advisor	Morgan Stanley
Anthropic data center contract	$50B

The $50 Billion Contract That Changed Everything

There's one reason above all others that explains Fluidstack's valuation surge.

A $50 billion custom data center construction agreement with Anthropic. That's a single contract -- one of the largest infrastructure deals in AI history.

Anthropic closed a $30B Series G in February 2026 and is approaching $30B in annualized revenue. A significant portion of that capital is being channeled into computing infrastructure, and Fluidstack has positioned itself as a core partner in that buildout.

The Investor Mix Is Telling

The co-leads on this round are worth paying attention to.

Jane Street is one of the world's largest quantitative trading firms. When a trading house -- not a traditional VC -- puts $1B into AI infrastructure, it signals the market sees this as a predictable returns play, not a speculative bet.

Situational Awareness is the fund run by Leopold Aschenbrenner, a former OpenAI researcher who made waves in 2024 with his essay arguing AGI could arrive sooner than most people expect. He's betting directly on the compute infrastructure that would make that possible.

The money isn't flowing to the companies building AI models. It's flowing to the companies building the places where AI models run. Classic picks-and-shovels strategy.

The Bigger Infrastructure War

This isn't just a Fluidstack story. The entire AI infrastructure market is on fire.

CoreWeave signed $35B in AI cloud contracts with Meta, Anthropic, and others in April 2026. A company that was pivoting from crypto mining just two years ago is now competing with AWS and Azure.

Of the $300B in total Q1 2026 venture investment, 80% ($242B) went to AI -- and infrastructure and compute captured the largest share within that.

Company	Focus	Recent Valuation
Fluidstack	AI data centers	$18B (in talks)
CoreWeave	AI cloud	$35B+
Lambda Labs	GPU cloud	$10B+
Crusoe Energy	Green AI infra	$5B+

What these companies share is clear: they don't build AI models. They build the physical spaces and power systems where AI runs. They look like software companies but their actual business is energy + real estate + hardware.

What This Means for You

The takeaway is straightforward.

The bottleneck in AI has fully shifted from models to infrastructure. It doesn't matter how good your model is if you can't find data center capacity to serve it. That's why Anthropic signs $50B infrastructure contracts and infrastructure company valuations jump 2.4x in four months.

For developers, two signals stand out. First, GPU cloud pricing isn't coming down anytime soon -- demand overwhelms supply. Second, local inference is becoming more valuable. With data center costs this high, running inference on local hardware like Intel's Arc Pro B70 becomes an increasingly attractive option.

The future of AI depends not just on who builds the smartest models, but on who can run them most efficiently.

References

• TechCrunch: AI data center startup Fluidstack in talks for $1B round at $18B valuation
• Bloomberg: Jane Street in Talks to Back Fluidstack at $18 Billion Valuation
• Sacra: Fluidstack revenue, funding & growth rate