CoreWeave posts $2.08B quarter, $740M loss

AI infrastructure is one of those businesses where “more demand” does not automatically mean “clean earnings.” That is basically the CoreWeave story right now. The company just posted a monster first quarter on revenue, but also a very large loss, and the stock dropped anyway because investors keyed in on the next quarter, not the last one. On May 7, CoreWeave reported $2.08 billion in Q1 2026 revenue and a $740 million net loss, then guided Q2 below what Wall Street had been looking for. ### What does CoreWeave actually sell? CoreWeave rents out AI compute — basically GPU-heavy cloud infrastructure for companies training and running large models. That puts it in the middle of the AI boom, serving labs, hyperscalers, and enterprises that need huge amounts of Nvidia-powered capacity fast. In Q1, the company said it signed new or expanded deals with Meta, Anthropic, Cohere, Jane Street, Mistral, and others. (investors.coreweave.com) ### So why did the stock fall? Because the market heard two things at once. First: demand is enormous. Second: the near-term math still looks rough. CoreWeave’s Q2 revenue outlook came in at $2.45 billion to $2.6 billion, with the midpoint below the analyst consensus around $2.69 billion. At the same time, it lifted the low end of its 2026 capital spending forecast to $31 billion from $30 billion, while keeping the top end at $35 billion. (investors.coreweave.com) ### Why is a $740 million loss not the whole story? Because this is a capital-hungry buildout business. CoreWeave is spending aggressively to add data centers, power, networking, and GPU capacity before revenue fully catches up. The company’s operating loss was $144 million, but interest expense alone was $536 million in the quarter. That matters, because it shows how much the financing structure is shaping the income statement. (cnbc.com) ### Where is the pressure coming from? From almost every expensive layer of the stack. Technology and infrastructure costs jumped to $1.27 billion in the quarter, and total operating expenses rose to $2.22 billion. CoreWeave also said component prices are rising, which is one reason it nudged capex guidance higher. Turns out AI demand is not just pulling revenue up — it is also making the hardware and buildout more expensive. (investors.coreweave.com) ### Then why are people still excited? Because the backlog is absurdly large. CoreWeave said revenue backlog reached $99.4 billion as of March 31, and that Q1 was the strongest bookings quarter in its history. It also said it has now surpassed 1 gigawatt of active power and has more than 3.5 gigawatts of total contracted power. In plain English — customers are reserving a lot of future capacity. (investors.coreweave.com) ### Is backlog the same as money in the bank? Not quite. Backlog is a strong signal of demand, but investors still have to believe CoreWeave can deliver that capacity on time, at acceptable cost, and with margins that improve instead of erode. That is the catch. A booked AI infrastructure business can still disappoint if build costs rise, financing gets heavier, or revenue lands later than expected. That is why a company can post huge revenue growth and still get punished. (investors.coreweave.com) ### How leveraged is this buildout? Very. CoreWeave said it raised $8.5 billion in new debt in the first quarter and ended the period with almost $25 billion in debt. Nvidia also bought another $2 billion in CoreWeave shares during the quarter, which underlines how strategically important the company has become in the AI supply chain. But debt-fueled expansion always makes investors more sensitive to any hint of slowing growth. (investors.coreweave.com) ### What is the real read-through? The market is not saying AI demand is fake. It is saying demand alone is not enough. CoreWeave looks like a company trying to become hyperscale at top speed, and the numbers show both sides of that bet — explosive customer demand, and a balance sheet doing heavy lifting to keep up. (cnbc.com) The bottom line is simple: CoreWeave’s quarter strengthened the bull case for AI infrastructure demand, but it also sharpened the bear case on cost, leverage, and execution. Investors are no longer arguing about whether customers want the capacity. They are arguing about what it will cost to deliver it. (investors.coreweave.com)