A Bitcoin miner just told Wall Street it expects to generate more than $4 billion in AI cloud revenue — and the market believed it. IREN’s shares surged approximately 15.7% on Monday after the company announced it had signed $2.8 billion in new multi-year cloud services contracts with AI developers, triggering a rapid upward revision to its annualized revenue target and sending a clear signal about where the real money in the mining sector is heading.
Summary
Key takeaways
- IREN raised its year-end AI cloud annualized run-rate revenue target to more than $4 billion after signing $2.8 billion in new multi-year contracts.
- About 85% of that revised target is already under contract with clients including Microsoft, Nvidia, Perplexity, and Figure AI.
- The company plans to scale AI cloud capacity from roughly 3 MW in 2025 to 480 megawatts in 2026, targeting 1.2 gigawatts by 2027.
- Customer prepayments cover approximately 45% of GPU capital expenditures, meaningfully reducing the company’s funding burden.
- IREN stock is the fourth-largest holding in the CoinShares Bitcoin Mining ETF (WGMI), which also rose 9.61% on the news.
IREN raises AI cloud revenue target beyond $4 billion
The numbers here are not incremental. Going from a prior target to one exceeding $4 billion on the back of a single contract announcement is the kind of revision that forces analysts to rethink the entire story around a company. For IREN, that story is no longer primarily about Bitcoin.
The $2.8 billion in new multi-year cloud services deals came from AI developers — a client base that has been aggressively hunting for reliable, high-performance computing capacity. IREN’s ability to land those contracts in volume suggests its infrastructure is being taken seriously at the enterprise level, not just as a niche crypto-adjacent play.
Market reaction and broader sector spillover
The stock’s roughly 15.7% single-day gain is striking, but the ripple effect matters too. The CoinShares Bitcoin Mining ETF (WGMI), where IREN is the fourth-largest holding at 10.05% of assets, climbed 9.61% as other miners also posted modest gains, according to Yahoo Finance. That kind of sector-wide lift reflects how investors are beginning to re-rate the entire Bitcoin mining cohort as a potential AI infrastructure play, not just a commodity-driven business.
Contract coverage and key partnerships with AI developers
With approximately 85% of the revised $4 billion target already under contract, IREN’s revenue outlook has a degree of visibility that most growth-stage companies cannot offer. The client roster explains why: Microsoft, Nvidia, Perplexity, and Figure AI are among those under agreement, spanning hyperscalers, chip architects, and frontier AI developers.
That combination is analytically important. Nvidia’s presence in the client list is particularly notable — it positions IREN not merely as a passive infrastructure host but as an active node in the GPU supply chain ecosystem. Co-CEO Daniel Roberts framed the demand environment directly: hyperscalers, enterprises, and AI developers are collectively seeking more capacity than IREN currently has available or has even planned to build.
What 85% contract coverage actually means
For investors, a target that is 85% contracted is a materially different proposition than a forward projection built on pipeline assumptions. It narrows the gap between ambition and execution risk. The remaining 15% leaves room for upside if new deals close, while the contracted base provides a foundation that is harder to dismiss as speculative.
Capacity expansion plans and funding strategy
The scale of what IREN is committing to is almost difficult to absorb in comparison to where it started. The company delivered roughly 3 megawatts of AI cloud capacity in 2025. By 2026, Co-CEO Daniel Roberts says it will deliver 480 megawatts — a roughly 160-fold increase in a single year. The 2027 target of 1.2 gigawatts would, if achieved, place IREN among the more substantial independent AI cloud infrastructure providers operating at scale.
Customer prepayments mitigate GPU capital expenditure needs
Funding that kind of buildout is the central execution challenge, and IREN has structured its contracts to partially address it. Recent customer agreements include prepayments that cover approximately 45% of the associated GPU capital expenditures. That means clients are effectively co-financing the infrastructure IREN needs to serve them — a model that reduces the company’s external funding dependency and aligns incentives on both sides of the deal.
It is a structurally smart arrangement. Rather than raising equity or debt at scale to fund speculative capacity, IREN is using contracted demand to offset a near-majority of its hardware costs upfront. The practical implication: the growth plan is partly self-funding through the very customers it is being built for.
Demand surpasses available AI cloud capacity as the business pivots
Perhaps the most analytically significant detail in IREN’s update is also the simplest: demand currently exceeds both available and planned capacity. That is not a supply-demand imbalance at the margin — it means IREN could theoretically commit to more infrastructure than it has announced and still find takers.
Repurposing Bitcoin mining infrastructure for AI computing
The vehicle for this expansion is a strategic repurposing of existing mining infrastructure. IREN is converting assets built for Bitcoin proof-of-work into high-performance computing facilities suited for AI workloads — a transition that research firm Bernstein flagged as early as April, when it assessed that IREN’s AI cloud business was on track to become its primary revenue driver. That call now looks well-supported by the contract figures.
This is part of a broader pattern playing out across the Bitcoin mining sector. Miners who built out significant power and data center capacity during the last crypto cycle now find themselves holding infrastructure that AI developers urgently need. The competitive advantage is not technical sophistication in AI — it is physical: land, power agreements, and cooling infrastructure that takes years to replicate from scratch. IREN’s pivot suggests it understood that advantage earlier than most and moved to monetize it through long-term contracted relationships rather than spot-market opportunism.
The question that follows from all of this is not whether IREN’s AI transition is real — the contract figures and client names make that case clearly enough. The more pressing question is whether the company can execute the 160-fold capacity expansion it has promised in a single year, in an environment where GPU supply chains, power permitting, and data center construction timelines rarely move as fast as revenue targets do.
FAQ
What is IREN’s updated AI cloud revenue target for the year-end?
IREN raised its year-end AI cloud annualized run-rate revenue target to more than $4 billion, following $2.8 billion in new multi-year cloud services contracts with AI developers.
Which major clients have IREN signed contracts with for AI cloud services?
Major clients currently under contract include Microsoft, Nvidia, Perplexity, and Figure AI, collectively accounting for approximately 85% of IREN’s revised revenue target.
How much AI cloud capacity does IREN plan to deliver in the coming years?
IREN plans to deliver 480 megawatts of AI cloud capacity in 2026, up from approximately 3 MW in 2025, with a target of 1.2 gigawatts by 2027.
How is IREN funding its GPU capital expenditures for AI cloud expansion?
Customer prepayments built into recent agreements cover approximately 45% of the associated GPU capital expenditures, substantially reducing IREN’s external funding requirements for those deployments.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.
