OpenAI Losses Increased Nearly 8X in 2025, with Spending Hitting $34B

I’m not sure how people are looking at numbers that show, even if we wipe off the enormous R&D expenditures, they are still in the red for inference + sales/marketing + admin and responding “this seems positive”.

It’s like being a sold a car and being told “well if you ignore the fact it has no engine it’s a good buy” yet it also has no wheels.

> Unless there's an assumption that R&D costs have to forever go up in order to increase revenue, I feel like this shows that the AI industry is actually on a path to profitability in the long term.

There are three futures right, I’ll rank them in order of fantasy -

1. Someone achieves AGI. At that point the economics of an individual company don’t even matter.

2. R&D costs do have to forever continue, because LLMs can be continually iteratively improved. Much like chip development, there is no end in sight, at least not on a near term timescale. If you are not continually at the frontier, customers will use a competitor or open/local alternatives.

3. LLMs reach a plateau of functionality. Further gains are minimal, quality reaches the apex of what the technology permits. In this scenario the hyperscalers have no business because open/local models will rapidly reach that same plateau as well.

What is counted as R&D is completely arbitrary. These figures are just playing accounting games to attempt to hide the massive ongoing costs.

We’ll see a little better when they IPO and are forced to attempt to make money but I wouldn’t invest in this business.

HSBC say they need to turn a 13b revenue to 200b by 2030 AND also find another 204b, in order to become profitable.

> Cost of Revenue: $7.5 billion

It's almost too good to be true. Did OpenAI intentionally leak this? It singlehanded eliminate the biggest concern: that tokens are sold at loss.

It's not going to happen.

Ceteris paribus, those figures imply a $45bn loss this year, $90bn loss next year and $110bn loss in 2028 before breakeven in 2029.

That's $250bn of losses to be financed from 2026 onwards. (They raised ~$120bn, $25bn up front and the rest based on milestones. So Another ~$125bn uncovered.) That only works if OpenAI stays a fundraising darling. So not a doomsday sceanario. But perilous, and dependent on short-term trends extending into long-term curves.

> Operating loss went from ~$8.8B to ~$20.9B — roughly 2.4x.

> Doesn't seem like a domesday scenario.

Those two lines are moving up and to the right, but are not parallel.

It all depends on where those two lines meet (the break-even point): too far in the future and the company will be dead anyway. Almost all companies will eventually be profitable; the problem is that the majority of them will need constant cash injections to keep the lights on.

Like the old aviation saying: even a brick will fly if it has enough thrust. doesn't make the brick a plane, though.

https://www.reuters.com/technology/openai-considers-drastic-...

I do wonder if this comparison is really meaningful. It looks like if they can grow infinitely, then at some point they should be profitable. However, that's already a somewhat sad story ("in the limit as x->inf, we'll actually _make_ money!"). And there are of course limitations. Anthropic, Google, open models etc are all real competitors, and it seems to me that there will only be one winner. If openAI is losing money faster than the others, then it may not survive long enough to reach that eventual profitability. And finally, the human population is limited. There isn't a true infinity that the pattern can extend to. If we've only reached 10% of the TAM that's fine, but if we're at like 70% (which personally I suspect is about right), then this looks bad.

Ed's claim is that they haven't shown inference to be profitable. Which is true? And that he personally believe it is unprofitable (his personal opinion, not what his data report).

I think that's a meaningful distinction with your statement

I'm not sure where they'd get that idea from? If inference was fundamentally unprofitable, I don't think we'd have seen the massive CapEx spend & VC cash flooding into AI, it'd be a negative gross margin trap if that were the case.

It looks unprofitable because of the massive CapEx spend right now to build data centers.

People that think inference is not profitable are mistaking the total compute cost as inference cost, when really it needs separated into training compute vs. inference compute.

The bigger question is, is when does training slow down, if at all? If we hit plateaus with LLMs, at that point inference becomes nearly pure profit once you own the compute (and a hardware refresh cycle every 3-5 years).

LLMs eventually hitting a dead end for more advanced capabilities is what would spell trouble for the labs. Any existing hyperscaler cloud can run inference all day long, as long as they have access to a model. They don't need OpenAI or Anthropic for that. The frontier labs entire valuations rely purely on them staying ahead of the commodity curve. The moment they can't do that, they're done.

No idea why his shit keeps getting submitted.