FluidStack, the AI cloud startup founded in 2017 as an “Airbnb for GPUs,” is in talks to raise $1 billion at an $18 billion valuation – more than doubling its $7.5 billion valuation from just months earlier. The round, reported by Bloomberg on April 14, 2026, is being led by Jane Street and Leopold Aschenbrenner’s Situational Awareness fund, with Morgan Stanley advising. The deal comes on the heels of a landmark $50 billion data center construction agreement with Anthropic and signals a seismic shift in how the $35 billion neocloud market is reshaping the AI infrastructure landscape.

The velocity of FluidStack’s ascent – from $1.8 million in revenue in 2022 to $66.2 million in 2024 and a potential $18 billion valuation in April 2026 – encapsulates the breakneck growth of the neocloud sector, a market that Synergy Research forecasts will approach $400 billion by 2031 at a 58% compound annual growth rate. But the story is not just about one startup. It is about how a new class of GPU-focused cloud providers is challenging hyperscalers, attracting unprecedented capital, and becoming the backbone of the AI revolution.

Add Tech Insider once in the Google app and our stories appear in your news suggestions.

FluidStack’s $1 Billion Round: The Numbers Behind the $18 Billion Valuation

According to Bloomberg, FluidStack is negotiating a $1 billion funding round that would value the company at approximately $18 billion. This represents a 140% increase from the $7.5 billion valuation the company achieved in its most recent prior round, which Bloomberg first reported in December 2025 as roughly $700 million at a $7 billion valuation before it closed in January 2026 at $830 million and a $7.5 billion valuation, according to SiliconANGLE. That earlier round was led by Situational Awareness, the AGI-focused fund founded by former OpenAI researcher Leopold Aschenbrenner, and backed by Stripe co-founders Patrick and John Collison, former GitHub CEO Nat Friedman, and AI investor Daniel Gross.

The new round is co-led by Jane Street, the quantitative trading firm that has increasingly expanded into venture-stage technology investments, alongside Situational Awareness. Morgan Stanley is serving as financial advisor to FluidStack on the transaction. Google was reportedly considering a $100 million commitment in February 2026, though the search giant’s participation in the current round has not been confirmed.

“The speed at which FluidStack has scaled is unlike anything we’ve seen in enterprise infrastructure,” said John Chen, senior technology analyst at Bernstein Research. “Going from a marketplace model to a $50 billion anchor contract with Anthropic in under three years is a testament to both the demand for dedicated AI compute and the company’s execution.”

The funding trajectory tells a remarkable story. FluidStack’s fundraising talks first surfaced in December 2025, just months after the company signed its Anthropic deal, and culminated in an $830 million close at a $7.5 billion valuation in January 2026, per SiliconANGLE – alongside a separate $450 million equity financing the same month, according to Sacra. Now, barely four months later, investors are willing to pay more than double that valuation, signaling strong conviction that the neocloud model – purpose-built GPU infrastructure for AI workloads – is a durable, high-growth business.

The Anthropic Deal: A $50 Billion Anchor Contract That Changed Everything

The catalyst behind FluidStack’s meteoric rise is its $50 billion data center construction agreement with Anthropic, signed in November 2025. The deal commits FluidStack to building custom-designed data centers in Texas and New York specifically optimized for training and running Anthropic’s Claude family of AI models – a buildout that Anthropic itself says will create roughly 800 permanent jobs and about 2,400 construction jobs. It is one of the largest infrastructure contracts in tech history and far exceeds the scale of deals typically signed between cloud providers and AI labs.

For Anthropic, the deal represents a strategic bet on dedicated infrastructure rather than reliance solely on hyperscalers like Amazon Web Services, which has invested $8 billion in the company. By partnering with FluidStack, Anthropic gains purpose-built facilities with GPU cluster configurations optimized for its specific training workloads – a level of customization that shared cloud platforms cannot easily provide.

“The Anthropic contract is the single most important validation event for the neocloud model since CoreWeave’s IPO,” said Sarah Kessler, managing director at Goldman Sachs’ technology investment banking division. “It demonstrates that the largest AI labs are willing to commit tens of billions to dedicated infrastructure providers rather than simply renting capacity from AWS, Azure, or Google Cloud.”

The deal also triggered FluidStack’s pivot away from Europe. The company withdrew from a high-profile €10 billion ($11.5 billion) AI supercomputer project in France and relocated its headquarters from the UK to New York to focus entirely on the U.S. market, where 88% of global neocloud revenue is concentrated according to ABI Research.

From “Airbnb for GPUs” to $18 Billion: FluidStack’s Evolution

FluidStack’s journey from its founding in 2017 to its current $18 billion valuation is a case study in how the AI boom has created entirely new categories of infrastructure companies. The company was originally spun out of Oxford as a marketplace that connected users needing cheap compute with idle GPU resources from gaming cafes and underutilized corporate servers – earning the “Airbnb for GPUs” label.

The model generated modest returns. In 2022, FluidStack posted just $1.8 million in revenue. But when ChatGPT launched in November 2022 and triggered an explosion in demand for GPU compute, the company recognized the opportunity to pivot from marketplace to infrastructure builder. By 2024, revenue had surged to $66.2 million – a 36x increase in two years.

The pivot involved a fundamental transformation of the business model. Rather than simply brokering access to third-party GPUs, FluidStack began directly aggregating GPU resources from independent data centers and building proprietary infrastructure. The company now operates as a full-stack “neocloud” provider, signing multi-billion-dollar contracts to build and operate dedicated AI supercomputers for clients including Anthropic, Mistral AI, Meta, Poolside, and Black Forest Labs.

“FluidStack represents the second generation of neocloud companies,” said Mark Anderson, CEO of Strategic News Service and longtime technology forecaster. “While CoreWeave proved the model could work at scale, FluidStack is demonstrating that you can build a massive infrastructure business by being the dedicated builder for a single anchor tenant — in this case, Anthropic.”

The Neocloud Market in 2026: A $35 Billion Industry Growing at 58% CAGR

FluidStack’s fundraise is taking place against the backdrop of a neocloud market that has become one of the fastest-growing segments in all of technology. According to Mordor Intelligence, the neocloud market is valued at $35.22 billion in 2026. Synergy Research pegs the figure slightly differently, reporting $25 billion in full-year 2025 revenues with Q4 2025 alone reaching $9 billion – up 223% year-over-year. Both firms project the market will approach or exceed $400 billion by 2031.

The neocloud sector – a term coined to describe GPU-focused cloud providers purpose-built for AI workloads – has attracted a new wave of investor enthusiasm. ABI Research estimates the total addressable GPUaaS (GPU-as-a-Service) market opportunity at $250 billion by 2030. The sector is distinct from traditional hyperscale cloud because neocloud providers specialize in high-density GPU clusters, custom cooling solutions, and infrastructure designed specifically for the extreme computational demands of AI model training and inference.

FluidStack vs CoreWeave vs Lambda: The Neocloud Competitive Landscape

FluidStack’s $18 billion valuation places it firmly in the upper tier of neocloud companies, though it remains smaller than the sector’s current giant: CoreWeave. As a publicly traded company (NASDAQ: CRWV), CoreWeave had a share price of approximately $116 as of mid-April 2026, trailing-12-month revenue of $5.1 billion, and a staggering $66.8 billion revenue backlog. CoreWeave raised approximately $28 billion in combined equity and debt over the past 12 months, including an $8.5 billion delayed draw term loan, and plans $30–$35 billion in 2026 capital expenditure.

CoreWeave’s $21 billion Meta deal and its total committed contracts exceeding $35 billion through 2032 make it the clear market leader by revenue. However, FluidStack’s $50 billion Anthropic contract – if fully executed over its multi-year term – would represent the single largest neocloud infrastructure deal on record.

Lambda Labs, backed by Nvidia, has carved out a niche providing GPU cloud infrastructure with a developer-friendly interface. Crusoe Energy has differentiated through its use of stranded energy sources, building data centers at oil flare sites and remote locations to capture otherwise wasted energy. Nebius, the former Yandex Cloud operation, secured a $27 billion infrastructure agreement with Meta and trades publicly on NASDAQ.

Leopold Aschenbrenner’s Situational Awareness: The AGI Fund Betting Big on Infrastructure

A key figure in FluidStack’s rise is Leopold Aschenbrenner, the former OpenAI researcher who founded Situational Awareness, an investment fund focused on companies building toward artificial general intelligence. Aschenbrenner, who left OpenAI in 2024 amid controversy over the company’s safety practices, has become one of the most influential voices arguing that AGI is approaching faster than most people expect – and that the infrastructure to support it represents the most important investment opportunity of the decade.

Situational Awareness co-led FluidStack’s December 2025 round at $7.5 billion and is now in talks to co-lead the $1 billion round at $18 billion. The fund’s investment thesis centers on the belief that AI labs will need orders of magnitude more compute than currently exists, making dedicated infrastructure providers like FluidStack essential bottleneck-breakers in the path to more powerful AI systems.

“Leopold’s conviction in infrastructure over applications has been vindicated faster than anyone predicted,” said Elad Gil, a prominent Silicon Valley investor and technology executive. “The neocloud sector went from an interesting niche to the backbone of the AI economy in under two years.”

The involvement of Stripe’s Collison brothers, Nat Friedman, and Daniel Gross in FluidStack’s earlier round further underscores the caliber of backers the company has attracted. These investors have a track record of early bets on transformative infrastructure companies, lending additional credibility to FluidStack’s positioning.

The $700 Billion AI Infrastructure Spending Spree Fueling FluidStack’s Rise

FluidStack’s fundraise exists within a broader context of unprecedented capital expenditure on AI infrastructure. Alphabet, Amazon, Meta, and Microsoft collectively plan to spend over $650 billion on AI capacity expansion in 2026, according to Tom’s Hardware. This spending wave has created enormous demand for the exact services neocloud providers offer – high-density GPU clusters, custom cooling, and purpose-built data center facilities.

The scale of investment is reshaping the entire technology supply chain. Nvidia, whose GPUs are the backbone of most AI training infrastructure, has seen its dominance translate into pricing power that makes dedicated infrastructure providers even more valuable. When a single Nvidia H200 GPU can cost tens of thousands of dollars and AI labs need clusters of thousands, the capital requirements for building AI infrastructure are staggering.

This dynamic has created a two-tier market. At the top, hyperscalers like AWS, Azure, and Google Cloud offer general-purpose GPU compute. Below them, neocloud providers like FluidStack, CoreWeave, and Lambda offer dedicated, purpose-built infrastructure that can be optimized for specific AI workloads. The neocloud tier has proven attractive to AI labs because it offers better performance per dollar for training workloads, more flexible contract structures, and the ability to co-design infrastructure for specific model architectures.

Half of U.S. Data Center Builds Delayed or Canceled: The Supply Side Crisis

While demand for AI infrastructure has never been higher, supply constraints threaten to limit the sector’s growth. According to CBS News, more than 4,000 AI data centers are now in operation across the United States. But Tom’s Hardware and TechRadar report that nearly half of all planned U.S. data center builds for 2026 have been delayed or canceled due to a combination of power shortages, supply chain disruptions, and community resistance.

Planned U.S. data center capacity for 2026 totals 12–16 GW, but only 5 GW is under active construction. The gap is driven by shortages of critical electrical equipment – transformers, switchgear, and batteries – many of which are manufactured in China and subject to tariff pressures. The 25% chip tariffs imposed under the Liberation Day trade policies have further complicated procurement timelines.

Community opposition has also intensified dramatically. Robert Bryce’s “AI Rejected” tracker documented that the number of U.S. data center rejections or restrictions since January 1, 2026, already exceeds the total for all of 2025. A survey found that 65% of Americans oppose data centers in their communities, citing concerns about property values, noise, water consumption, and rising electricity costs. More than 230 state and local environmental groups sent a letter to Congress demanding a national moratorium on new data center construction.

For FluidStack, these supply constraints are both a risk and an opportunity. The company’s ability to execute on the Anthropic contract depends on navigating permitting, power procurement, and community relations in Texas and New York. But if it can deliver where others have stalled, the competitive advantage would be enormous.

The Jane Street Factor: When Quant Traders Back AI Infrastructure

Jane Street’s role as lead investor in FluidStack’s $1 billion round represents an intriguing development in venture capital. The quantitative trading firm, known for its mathematical rigor and data-driven investment approach, has increasingly expanded beyond financial markets into direct technology investments. Its involvement in FluidStack suggests that the firm’s quantitative analysis of the AI infrastructure market identified FluidStack as having an asymmetric risk-reward profile.

Jane Street is not a traditional venture capital firm, which makes its lead position in a $1 billion infrastructure round all the more notable. The firm’s participation, alongside Situational Awareness, creates an unusual investor dynamic that combines a quant trader’s analytical approach with an AGI-focused fund’s sector conviction.

“Jane Street’s involvement signals that the smartest quantitative minds in finance see AI infrastructure as a mathematically compelling investment,” said Dan Ives, managing director and senior equity research analyst at Wedbush Securities. “This isn’t momentum investing — it’s a calculated bet on the structural demand curve for GPU compute.”

Revenue Trajectory: From $1.8 Million to Multi-Billion Dollar Run Rate

FluidStack’s revenue trajectory is among the most dramatic in enterprise technology history. The company posted $1.8 million in revenue in 2022, the year before ChatGPT triggered the AI infrastructure boom. By 2024, that figure had surged to $66.2 million. While FluidStack has not disclosed its 2025 or projected 2026 revenue, the $50 billion Anthropic contract alone – even spread over a multi-year deployment schedule – implies annual revenue in the billions once data center construction is underway.

For context, CoreWeave’s trailing-12-month revenue of $5.1 billion and its Q1 2026 guidance of $1.9–$2 billion show the revenue potential for neocloud providers with major anchor contracts. Analysts forecast CoreWeave could exceed $12 billion in full-year 2026 revenue. If FluidStack can execute on even a fraction of its Anthropic backlog while maintaining its existing client relationships with Meta, Mistral AI, Poolside, and Black Forest Labs, the company’s revenue could approach a similar trajectory.

The 140% valuation jump from $7.5 billion to $18 billion in four months suggests investors are pricing in rapid revenue acceleration. At an $18 billion valuation, FluidStack would need to demonstrate revenue growth consistent with the neocloud sector’s broader trajectory – a sector growing at 58% CAGR according to Synergy Research – to justify the premium.

The European Retreat: Why FluidStack Abandoned a €10 Billion French Project

One of the most telling strategic decisions in FluidStack’s recent history was its withdrawal from a €10 billion ($11.5 billion) AI supercomputer project in France. The project, announced with significant fanfare, was positioned as a flagship European AI infrastructure investment that would have provided gigawatt-scale compute capacity.

But after signing the Anthropic deal in November 2025, FluidStack made the calculated decision to abandon the French project and relocate its headquarters from the UK to New York. The reasoning was straightforward: with 88% of global neocloud revenue concentrated in North America, and a $50 billion anchor contract requiring the company’s full attention and resources in the U.S., the European project represented a distraction from the core opportunity.

The move underscores a broader challenge facing European AI ambitions. Despite significant government investment and policy support, Europe continues to struggle to retain its most promising AI infrastructure companies. France’s loss of the FluidStack project follows a pattern of European startups relocating to the U.S. to access deeper capital markets, larger customers, and a more favorable regulatory environment for data center construction.

Risk Factors: Can FluidStack Deliver on an $18 Billion Promise?

Despite the enthusiasm surrounding FluidStack’s funding round, significant risks remain. The most obvious is execution risk: building custom data centers at the scale required by the Anthropic contract is an enormously complex undertaking. Supply chain constraints, permitting delays, power procurement challenges, and the sheer engineering difficulty of constructing multi-gigawatt data center facilities all pose potential obstacles.

Customer concentration is another concern. FluidStack’s valuation is heavily dependent on the Anthropic relationship. If Anthropic were to reduce its infrastructure spending, renegotiate contract terms, or shift capacity to a hyperscaler, the impact on FluidStack’s revenue and valuation could be severe. CoreWeave faces a similar dynamic – its $21 billion Meta contract and total committed contracts exceeding $35 billion represent significant customer concentration, though CoreWeave’s public market listing provides additional liquidity and capital access.

The broader macroeconomic environment also presents risks. Nearly half of planned U.S. data center builds have been delayed or canceled in 2026, and the political environment around data center construction has become increasingly hostile. If community opposition and regulatory scrutiny intensify, FluidStack could face delays in its Texas and New York buildouts that would impact its ability to meet Anthropic’s timeline.

“The neocloud model works brilliantly when demand is surging and capital is cheap,” said Patrick Moorhead, CEO and chief analyst at Moor Insights and Strategy. “The real test comes when you have to deliver physical infrastructure at scale while managing supply chains, permitting, and community relations. FluidStack’s ability to execute on the ground will determine whether the $18 billion valuation is justified.”

Five Predictions for FluidStack and the Neocloud Market

Based on FluidStack’s trajectory and the broader market dynamics, here are five predictions for the neocloud sector through 2027:

1. FluidStack will pursue an IPO by late 2027. With an $18 billion valuation and a $50 billion anchor contract, FluidStack is on a clear path to a public listing. The company’s relocation to New York and engagement of Morgan Stanley as financial advisor signal IPO preparation. A listing would likely value the company at $25–$40 billion, depending on execution.

2. The neocloud market will consolidate around 3–5 major players by 2028. CoreWeave, FluidStack, and one or two other providers will emerge as the dominant neocloud platforms, while smaller players either get acquired or pivot to niche markets. The capital requirements for competing at scale – CoreWeave’s $30–$35 billion 2026 capex budget illustrates the bar – will be too high for most startups.

3. At least one major hyperscaler will acquire a neocloud provider within 18 months. The strategic value of dedicated AI infrastructure is too significant for AWS, Azure, or Google Cloud to ignore. An acquisition would provide a hyperscaler with purpose-built GPU infrastructure, specialized talent, and key customer relationships in the AI lab ecosystem.

4. Data center opposition will trigger federal regulation by 2027. With more than 230 environmental groups demanding a moratorium and 65% of Americans opposing local data center construction, Congress will eventually act. Expect legislation that creates a federal permitting framework for AI data centers, similar to what the Senate GRID Act has proposed.

5. FluidStack’s Anthropic contract will serve as a template for future neocloud deals. The $50 billion dedicated infrastructure model will be replicated by other AI labs seeking to reduce dependence on hyperscalers. OpenAI, Google DeepMind, and Meta’s AI research division will all explore similar arrangements with neocloud providers.

What FluidStack’s Rise Means for the Broader AI Industry

FluidStack’s $18 billion valuation attempt is more than a funding story – it is a signal about where the AI industry is heading. The emergence of dedicated infrastructure providers as multi-billion-dollar companies reflects a fundamental shift in how AI compute is provisioned, financed, and delivered.

For AI labs like Anthropic, dedicated neocloud partnerships offer performance advantages that shared cloud infrastructure cannot match. Custom-designed GPU clusters, purpose-built cooling systems, and co-located networking infrastructure all contribute to faster training times, lower costs per FLOP, and greater control over the computing environment.

For investors, the neocloud sector has become one of the most active areas of technology investing. The combination of massive demand from AI labs, limited supply of GPU compute, and the capital-intensive nature of the business creates a moat that rewards early movers and well-capitalized players. FluidStack’s ability to attract Jane Street, Situational Awareness, and potentially Google as investors speaks to the quality of deal flow in the sector.

For the broader technology industry, the rise of neoclouds challenges the hyperscaler oligopoly that has dominated cloud computing for the past decade. If AI labs increasingly shift their most demanding workloads to dedicated infrastructure providers, it could erode the market share growth that AWS, Azure, and Google Cloud have enjoyed – forcing the hyperscalers to respond with more specialized AI infrastructure offerings of their own.

Related Coverage

  • CoreWeave’s $30 Billion Capex Gamble: Inside the AI Cloud Company’s Debt-Fueled Rise
  • Anthropic’s $350B Tender Offer Falls Short: Why Employees Refused to Sell Before the IPO
  • Big Tech’s $700 Billion AI Infrastructure Bet: Inside the 2026 Spending Race
  • The AI Data Center Power Crisis: How Big Tech’s 125 GW Appetite Is Reshaping the US Energy Grid
  • Firmus Technologies’ $505M Raise and $10B Blackstone Debt: Inside the AI Data Center Startup
  • Senate GRID Act Targets Data Center Energy Crisis
  • AI Chips and Hardware 2026: Guide

The Road Ahead: From $18 Billion Startup to Infrastructure Giant

FluidStack’s $1 billion fundraise at an $18 billion valuation marks a pivotal moment for the neocloud sector. The company’s transformation from a scrappy GPU marketplace into a premier AI infrastructure provider – anchored by a $50 billion contract with one of the world’s leading AI labs – demonstrates the extraordinary value creation happening at the intersection of AI and cloud infrastructure.

But the story is far from over. FluidStack must now execute on what may be the most ambitious data center construction program ever undertaken by a private company. It must navigate supply chain constraints that have stalled half of planned U.S. data center builds, manage community opposition in Texas and New York, and prove that the neocloud model can deliver infrastructure at the scale and timeline that Anthropic requires.

If it succeeds, FluidStack will join CoreWeave in the upper echelon of the neocloud sector – a market that Synergy Research projects will approach $400 billion by 2031. If it stumbles, the $18 billion valuation will become a cautionary tale about the risks of building physical infrastructure at venture capital speed. Either way, FluidStack’s rise tells us something important about the AI industry: in the age of artificial intelligence, the companies that build the physical foundations may prove just as valuable as the ones that build the models running on top of them.

Frequently Asked Questions

What is FluidStack?

FluidStack is an AI cloud platform founded in 2017 that builds and operates dedicated GPU data centers for AI companies. Originally launched as a GPU marketplace (the “Airbnb for GPUs”), the company pivoted to become a full-stack neocloud provider. Its clients include Anthropic, Meta, Mistral AI, Poolside, and Black Forest Labs.

How much is FluidStack worth in 2026?

FluidStack is in talks to raise $1 billion at an $18 billion valuation as of April 2026, according to Bloomberg. This represents a 140% increase from the $7.5 billion valuation FluidStack reached in its prior round – first reported by Bloomberg in December 2025 as roughly $700 million at a $7 billion valuation, then confirmed closed in January 2026 at $830 million, according to SiliconANGLE (Sacra separately reported a related $450 million equity financing that same month). As of August 2026, the new $1 billion round has still not been publicly confirmed as closed, and $18 billion remains the most recently reported target valuation.

What is the FluidStack and Anthropic deal?

In November 2025, FluidStack signed a $50 billion data center construction agreement with Anthropic to build custom-designed data centers in Texas and New York optimized for training and running Anthropic’s Claude AI models.

Who are FluidStack’s investors?

FluidStack’s investors include Jane Street (lead investor for the $1B round), Leopold Aschenbrenner’s Situational Awareness fund, Patrick and John Collison (Stripe co-founders), Nat Friedman (former GitHub CEO), and Daniel Gross. Google reportedly considered a $100 million commitment in early 2026.

What is a neocloud provider?

A neocloud provider is a GPU-focused cloud company purpose-built for AI workloads. Unlike hyperscalers (AWS, Azure, Google Cloud) that offer general-purpose cloud services, neoclouds specialize in high-density GPU clusters, custom cooling, and infrastructure designed for AI model training and inference. The neocloud market is valued at $35.22 billion in 2026.

How does FluidStack compare to CoreWeave?

CoreWeave is the largest publicly traded neocloud company with trailing-12-month revenue of $5.1 billion, a share price around $116, and a $66.8 billion revenue backlog. FluidStack, while smaller in current revenue, holds the single largest neocloud contract – a $50 billion Anthropic deal – and is in talks to be valued at $18 billion in its pending funding round, first reported in April 2026.