Home / Research / Neocloud Market Statistics

Neocloud Market Statistics: 2026 Report

By Darin ten Bruggencate, Founder, AI Industry Reviews · September 17, 2026

Training and running AI models takes huge numbers of GPUs. A group of specialist companies called neoclouds buys those chips in bulk, installs them in data centers and rents them to AI developers by the hour or on multi-year contracts. They have grown fast, but a few customers provide most of their revenue, so we looked at how big they are and how they are financed.

Our analysts tracked 23 of these providers using quarterly filings, shareholder letters, published rate cards, and announced power contracts. Where a company is private, we built figures from disclosed funding rounds, announced customer contracts, and capacity filings with utilities and planning authorities.

Leading Neoclouds by Revenue and Contracted Power

The table ranks the seven largest specialist providers by annualized revenue from third-quarter 2026 results, with the power they have energized and the power they have under contract. Contracted power is the better forward indicator, since a gigawatt under contract converts to revenue only as it comes online.

The Leading Neoclouds by Revenue and Contracted Power, 2026

Insights

• We found contracted power to exceed active power by a factor of 5.0 across the category.
• Crusoe and Nebius each hold more contracted power than CoreWeave while generating less than a third of its revenue, a gap that reflects build timelines.
• Seven providers now top half a billion dollars in annualized revenue.

Neocloud Quarterly Revenue, Q1 2024 to Q3 2026

The category's growth is easier to read quarter by quarter than in annual totals, because the largest contracts step up in jumps as capacity energizes. In the table below, we separate CoreWeave and Nebius from the rest of the field, then plot all three series.

The Neocloud Quarterly Revenue, 2026

CoreWeave's lead has held in percentage terms while widening in absolute dollars, and the rest of the field, taken together, has grown faster than CoreWeave since the middle of 2025. Second and third tier providers are now winning contracts that would previously have gone to whoever had capacity first.

Customer Concentration and Contract Structure

A small number of customers account for most of the revenue in this category, and those customers are also building their own capacity. In the table below, we report concentration and contract structure for the five providers with sufficient disclosure to measure it.

The Customer Concentration and Contract Structure, 2026

Concentration of this size is normally a warning sign, but here it also supports the financing: a four-year take-or-pay contract with a creditworthy counterparty lets these companies raise the debt that buys the GPUs in the first place. The larger exposure arrives when the contract ends. The customer has by then built its own capacity, and the provider is left holding three-year-old accelerators in a market where the replacement generation is twice as efficient. The providers with the shortest contracts and the least concentration have the lowest exposure to that scenario, and most of the category has not made that trade.

Neocloud Unit Economics and Leverage

In the table below, we compare the accounting assumptions and balance sheets behind the revenue figures above. The depreciation schedule has the largest effect of any line: every year added to the assumed useful life of a GPU moves cost out of the current period and into a future one.

The Neocloud Unit Economics and Leverage, 2026

Insights

• We found a two-year spread in assumed GPU useful life across the category, which on a $30,000 accelerator is a difference of roughly $2,500 per card per year in reported cost.
• Adjusted EBITDA margins cluster between 29 and 58 percent while every provider we measured reported a net loss, and interest and depreciation account for the gap.
• The category carries about $36 billion in GPU-backed debt, so lenders and shareholders are exposed if GPU resale values fall faster than expected.

Neocloud Share of the AI Cloud Market Through 2030

In the table below, we set our estimate of total AI cloud spend against the share captured by specialist providers, and against the inference share of neocloud workloads.

The Neocloud Share of the AI Cloud Market, 2026

The share figure peaks before the end of the forecast period. Specialist providers win share while capacity is scarce and the buyers are model developers who care mainly about interconnect quality. As the workload mix shifts toward inference, the buyer changes: an enterprise serving a product wants the compute next to its data, its identity system, and its compliance controls, which is the hyperscalers' home ground. Our projection assumes the neoclouds answer that by moving up the stack into managed inference, and the providers already doing so carry the shortest contracts and the least debt.

What Neocloud Market Trends Mean for 2027 and Beyond

Newer GPU cloud companies have become a real alternative to the big cloud providers for renting AI computing power. They are growing fast and earn healthy operating income, but none of them is profitable yet, and much of their growth is paid for with debt backed by the chips themselves.

Their future depends on finishing data centers on schedule and on GPUs keeping their value. Delays or a drop in used GPU prices would hit the most indebted providers hardest.

So if you rent capacity from a newer GPU cloud, check its financial health and keep a backup provider, because a few of these companies carry heavy debt.

Requesting a Copy of This Report

If you would like a PDF copy of this report, or the provider-level dataset behind it, you can reach out here.