This Artificial Intelligence (AI) Stock Has a $66 Billion Backlog. Here's Why It Could Soar in 2026.

Artificial intelligence (AI) is expensive. It constantly needs more of everything to advance as a technology: more electricity, more data centers, more memory, more money.

Back in 2024, OpenAI CEO Sam Altman stated that ChatGPT-4, then the latest version of the large language model (LLM), cost over $100 million to develop.

Lots of companies want to develop AI, either for commercial use or for their own internal purposes. But many either can’t spend or would rather avoid spending hundreds of millions, if not billions, of dollars to develop their own AI infrastructure.

Fortunately for them, there’s **CoreWeave **(CRWV +8.12%), which offers prospective AI developers access to the tools and hardware needed to develop AI without having to take on all the costs associated with it themselves.

Image source: Getty Images

Rent-a-lab

CoreWeave offers a line of products aimed at AI developers. For instance, CoreWeave ARENA, which is basically a rentable AI lab on the company’s cloud.

In it, customers can test their AI program’s performance, see how effectively it can scale, and how much it will cost. It runs on CoreWeave’s own data center network, which spans North America and Europe. With access to CoreWeave’s systems, start-up AI developers don’t have to shell out loads of money to build their own data centers.

The company also offers a service called Slurm on Kubernetes or SUNK, which is an industry first. It allows customers to train their AI programs with CoreWeave’s network. That’s important because training new AI models wasn’t cheap when OpenAI launched the first version of ChatGPT, and it’s only becoming more expensive.

According to an Epoch AI study, the cost of training new AI models has been growing at a rate of 2 to 3 times annually for the past eight years. And if that rate continues, it could cost upwards of $1 billion just to train an LLM by 2027.

But with CoreWeave, smaller AI companies can train their AI faster and cheaper than the competition in some cases. Mistral AI, a French AI developer, trained its eponymous AI program fully on CoreWeave’s cloud and was able to do it 2.5 times faster than it would have been able to manage on its own.

And it’s not just smaller companies using CoreWeave’s services either. International Business Machines is also a major customer, and it partnered with CoreWeave to develop and train its Granite models 1.8 times faster than it could prior to using CoreWeave’s services.

OpenAI also uses CoreWeave to aid in the creation of its own models. It makes sense; they know firsthand how pricey developing AI is.

How does that translate to CoreWeave’s financial results? Quite well, actually.

Expand

NASDAQ: CRWV

CoreWeave

Today’s Change

(8.12%) $6.08

Current Price

$81.00

Key Data Points

Market Cap

$39B

Day’s Range

$77.90 - $82.97

52wk Range

$33.52 - $187.00

Volume

1.1M

Avg Vol

28M

Gross Margin

47.77%

An AI growth machine

CoreWeave had a pretty incredible end to 2025.

While profitability eluded the company with its $606 million adjusted net loss, its revenue grew 168% over 2024 to hit $5.1 billion. And new business is flooding in.

The company grew its backlog 342% to $66.8 billion by the end of 2025. So, its services are clearly in very high demand.

Over half of that backlog is in long-term deals of 25 months or more, so revenue is set to continue coming in at a good clip moving forward.

CoreWeave is anticipating an even better 2026 and has set its full-year guidance at $12 to $13 billion. The low-ball for guidance is more than double its total revenue for 2025.

A tad optimistic? Maybe.

But even with fears of an AI bubble on the horizon spooking investors throughout 2025, CoreWeave still managed to grow at a remarkable pace. And with new deals coming in at an incredible rate, CoreWeave’s growth guidance is likely attainable.

There’s no shortage of demand, and CoreWeave has already locked in long-term partnerships with tech giants like IBM.

CoreWeave is already looking strong even while running at a loss. If it can achieve profitability in the relative short term, then it’s likely to remain as an important piece of the AI market’s infrastructure.

Give this one a look now, it’s down from its all-time high in June 2025, but still up 81% from where it was at its initial public offering (IPO) in March 2025. With growth numbers like the ones it has been releasing, CoreWeave seems to be hitting its stride.