Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Leaked OpenAI financials show $38.5B loss and compute burn

runtimewire.com

Post reply on HN
Print view View original

241–250 of 278 posts

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#241

Post by gspetr »

somenameforme wrote: post #76
lukeschlather wrote: post #12

The R&D expenditure seems reasonable, and the revenue numbers seem realistic. I have no trouble believing they can be profitable by 2030 or much sooner. What I don't get is how you get from $30B in revenue to a nearly $1T valuation, but that seems almost level-headed compared to SpaceX, and it's not like any of the big tech companies' valuations make much sense in the context of their revenue.

The market is pricing in the potential for future revolutionary shifts which seem fairly likely. For instance if there ends up being substantial labor disruption due to LLMs then the economy as we know it is going to end up being reshaped in ways that are difficult to imagine beyond the fact that the LLM providers would likely play a critical role in it. Similarly, SpaceX has already brought the cost of getting thing…

$20/kg is a wild claim.

A typical American semi-truck (18-wheeler) can carry between 40,000 and 45,000 pounds (20 to 22.5 tons) can travel only 160k miles (2/3 distance to the Moon) on a budget of $400k (i.e. $20/kg).

Do you expect us to believe that rocket transportation would be in the ballpark of truck prices in the near future?

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#242

Post by ltononro »

The headline is a little bit misleading and sensationalist imo Most of the loss comes from the non-profit->for-profit switch This is not operational loss, one-time stuff to make the company for-profit and compete with Anthropic in apples-to-apples comparison, am I right?

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#243

Post by ltononro »

jacobgold wrote: post #155
simonw wrote: post #19

"OpenAI generated $13.07 billion in revenue in 2025" Considering just four years ago they were a research lab with hardly any revenue at all, and no corporate muscles for earning revenue, I think that is a very impressive number. (Sure, they're losing a whole lot of money too. Same goes for almost every other hyper-growth company in the history of tech.)

Annual revenue of $13 billion per year puts them on par with Apple's AirPods revenue, which places them in U.S. Fortune 500.

lol so weird to put that into perspective macbook ~8b airpods ~18b iphones ~60b

I know they are massive, but AI seems something much more important than airpods

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#244

Post by vunderba »

Razengan wrote: post #193
echelon wrote: post #86

Earlier quoted context omitted.

Yeah, but they have no moat. They gave up on video because three separate Chinese companies were kicking their ass (and for cheaper). Google has a better image model in the majority of cases. Much faster, too. Claude Opus and Fable are like a billion times better. It's not even funny. Codex can't do Rust at all. What does that leave them? Ads in ChatGPT? I've started to just rely on Google search blended with Gemini…

> Google has a better image model in the majority of cases. Not always. A couple months ago (before ChatGPT Images 2) I tried various prompts on both Google's Nano Banana or whatever and ChatGPT. "Capybara riding a tricycle. It has 7 tentacles instead of legs" Google got the number of tentacles completely wrong: https://i.postimg.cc/nzY30y7X/Capybara-Gemini-Nano.png and after some additions like spotted fur and multi…

I use LLMs more in the context of peer-reviewing and also came to a similar conclusion, gpt-5.5 codex xhigh reasoning seemed to catch more edge cases and went "deeper" into analysis than Opus 4.7/4.8.

My preliminary tests of Fable were pretty promising but that's DOA for everyone for now.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#245

Post by JSR_FDED »

spongebobstoes wrote: post #82
JSR_FDED wrote: post #50

Why would revenue continue to grow at this rate? Enterprises are becoming increasingly aware that the best models can be used for planning and then cheaper models for execution - all the way to local models for some tasks. Add in increasing competition from Chinese models… I’m not convinced this revenue growth is guaranteed.

because they have more research and better models coming building a Rube Goldberg machine on Chinese models might work okay, but it will be brittle, and is unlikely to work as well as the latest and greatest model from OpenAI the demand for intelligence is nearly limitless

Why would it be a Rube Goldberg machine? You talk to the Chinese models through the exact same API as OpenAI. They do tool calling in the same way. They deliver 90% of the leading benchmarks at less than one tenth the cost.

There’s also no reason to think they’ll always be lagging the US. At some point the scarcity of GPUs will be resolved, or they’ll brute-force the problem using less capable chips but leveraging their much lower energy costs to make this viable.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#246

Post by JSR_FDED »

danielmarkbruce wrote: post #72
JSR_FDED wrote: post #50

Why would revenue continue to grow at this rate? Enterprises are becoming increasingly aware that the best models can be used for planning and then cheaper models for execution - all the way to local models for some tasks. Add in increasing competition from Chinese models… I’m not convinced this revenue growth is guaranteed.

global gdp is over 100 trillion. Something like 50-60% is paid to labor. If you assume AI takes a good chunk of labor, the market is gigantic. Really really really gigantic.

That’s a good way of looking at it, I appreciate the different perspective.

From your headline number you’d have to deduct non-knowledge work though.

You’d also have to take into account the fact that while AI can often replace some tasks, it’s often not enough to replace the entire worker.

For the high end knowledge worker jobs the corresponding token costs could be higher than the cost of wages.

Given the demand you describe, would it even make a difference whether you invested in OpenAI or Anthropic?

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#247

Post by etempleton »

raincole wrote: post #146
cmiles8 wrote: post #85

OpenAI likely missed the window to have a successful IPO. A year ago, even 6 months ago, folks would have been still hypnotized by the hype and they would have pulled it off. Today too many people see a burning ship of cash and no moat to justify the burn. The story just isn’t there anymore.

After reading Financial Times and Ed Zitron's articles[0][1], I've reached the opposite conclusion. OpenAI's situation healthier than what the outsiders once believed: > Revenue: $13.07 billion > Cost of Revenue: $7.5 billion In other words generating tokens is actually a profitable business even for the frontier models. It's best to IPO when it's the case. [0]: https://www.ft.com/content/e15b0d7e-ff6b-4f16-ba7a-4068…

Assuming that constant R&D is not a requirement to compete. At what point can any of these companies stop improving their models? The answer is when they have a monopoly position. And we’re nowhere near that happening.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#248

Post by Paradigma11 »

slow_typist wrote: post #69
bgirard wrote: post #43

Earlier quoted context omitted.

You're missing the point. There was a lot of debate around if inference was subsidized or not. And that's a huge point to confirm in the public discourse.

Sure, but being able to pay for inference and nothing but inference out of revenue leads to what end?

To them going bankrupt and users paying another company, that bought them cents to the dollar, the same money for the same product.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#249

Post by skybrian »

somenameforme wrote: post #225
skybrian wrote: post #213

Earlier quoted context omitted.

So what's the consensus? Any article in particular that you recommend?

Peruse the forum. There's a topic on near everything relevant you might imagine. If you want a topic starting from a bearish premise, here is one. [1] In general I think there's little doubt that Starship is viable, but the exact implications are open to a wide range of speculation. Prices are going to go down, and payload sizes are going to go up, but the exact degree is open to speculation varying by orders of magn…

These links seem to be blocked now. Edit: working again.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#250

Post by Razengan »

vunderba wrote: post #244
Razengan wrote: post #193

Earlier quoted context omitted.

> Google has a better image model in the majority of cases. Not always. A couple months ago (before ChatGPT Images 2) I tried various prompts on both Google's Nano Banana or whatever and ChatGPT. "Capybara riding a tricycle. It has 7 tentacles instead of legs" Google got the number of tentacles completely wrong: https://i.postimg.cc/nzY30y7X/Capybara-Gemini-Nano.png and after some additions like spotted fur and multi…

I use LLMs more in the context of peer-reviewing and also came to a similar conclusion, gpt-5.5 codex xhigh reasoning seemed to catch more edge cases and went "deeper" into analysis than Opus 4.7/4.8. My preliminary tests of Fable were pretty promising but that's DOA for everyone for now.

Claude often spent most of its output listing all the things that were already correct and working! "This is good"

and most of its findings were false positive or outright wrong as in the screenshot I posted above.

Post reply on HN