How high are OpenAI’s compute costs? Possibly a lot higher than we thought

Stay informed with free updates

Simply sign up to the Technology sector myFT Digest — delivered directly to your inbox.

Elsewhere on the internet today, tech blogger Ed Zitron has an interesting post about OpenAI’s cash burn. The gist is that OpenAI’s running costs may be a lot more than previously thought, and that its main backer Microsoft’s doing very nicely out of their revenue share agreement.

The post cites data purportedly showing OpenAI’s inference spend on Microsoft’s Azure web-hosting platform. Inference is the process by which applications such as ChatGPT call on large-language models to generate responses.

Pre-publication, Ed was kind enough to discuss with us the information he has seen. Here are the inference costs as a chart:

The best place to begin is by saying what the numbers don’t show. The above is understood to be for inference only, not the more intensive task of model training, and it can’t give a full picture of Microsoft’s multi-strand relationship with OpenAI.

More importantly, is the data correct? We showed Microsoft and OpenAI versions of the figures presented above, rounded to a multiple, and asked if they recognised them to be broadly accurate. We also put the data to people familiar with the companies and asked for any guidance they could offer.

A Microsoft spokeswoman told us: “We won’t get into specifics, but I can say the numbers aren’t quite right.” Asked what exactly that meant, the spokeswoman said Microsoft would not comment and did not respond to our subsequent requests. An OpenAI spokesman did not respond to our emails other than to say we should ask Microsoft.

A person familiar with OpenAI said the figures we had shown them did not give a complete picture, but declined to say more. In short, though we’ve been unable to verify the data’s accuracy, we’ve been given no reason to doubt it substantially either. Make of that what you will.

Taking everything at face value, the figures appear to show a disconnect between what’s been reported about OpenAI’s finances and the running costs that are going through Microsoft’s books.

For example, in the first six months of 2025, OpenAI appears to have spent close to $5bn just on inference. Its cash burn over the same period was reported to be $2.5bn on revenue of $4.3bn.

The information reported by Ed also includes what’s purported to be quarterly figures for Microsoft’s 20 per cent revenue share. These need extra caution.

As well as being due 20 per cent of OpenAI’s revenue, according to reports, Microsoft gives OpenAI a 20 per cent cut of certain Azure and Bing revenues. Royalty payments might also apply. It’s a complicated, somewhat circular non-public relationship that can’t be summed up in a single data point.

By multiplying the revenue share number by five, we get group revenue estimates for OpenAI. They’re lowball estimates by definition, because of the two-way nature of the companies’ relationship, but they could still offer an insight into how revenue and running costs have scaled together.

Here’s the chart:

As Ed writes, OpenAI appears to have spent more than $12.4bn at Azure on inference compute alone in the last seven calendar quarters. Its implied revenue for the period was a minimum of $6.8bn. Even allowing for some fudging between annualised run rates and period-end totals, the apparent gap between revenues and running costs is a lot more than has been reported previously. And, like Ed, we’re struggling to explain how the numbers can be so far apart.

If the data is accurate — which we can’t guarantee, to reiterate, but we’re writing this post after giving both companies every opportunity to tell us that it isn’t — then it would call into question the business model of OpenAI and nearly every other general-purpose LLM vendor. At some point, going by the figures, either running costs have to collapse or customer charges have to rise dramatically. There’s no hint of either trend taking hold yet.

And what happens if nothing changes? Well, it’s a silly question for obvious reasons, but we’ll answer it anyway.

Using the rolling four-quarter growth rates, based on the data as presented, OpenAI’s implied group minimum revenue before its sharing agreements might be covering the inference costs by around 2033. Net of the 20 per cent Microsoft revenue share, OpenAI’s minimum estimated revenue would cover inference costs approximately never: