Which is such a wild thing about our industry. These can be simultaneously true
- No one has available tokens/gpus/datacenter for sale at any price
- No one will lend people money to build it
- No one has available tokens/gpus/datacenter for sale at any price
- No one will lend people money to build it
But the cost to add more power capacity is huge and highly variable. Just because oil is expensive today doesn’t mean the capex to pump more is justified.
Really what AI needs is large customers that will commit to large spend over the next 5-10 years. Without this AI companies are basically offering duration arbitrage of electricity. Not a good place to be; the finances will always look bad.
Unfortunately, the amount of capital required by the entire AI industry is enough that counterparty risk is now a concern.
Everyone is using more and more tokens, everyone is complaining that providers lobotomize flag ship models because they don't have the capacity to serve everyone.
Then they turn around and compare it to the dark fiber bust or railways or whatever when in each of those the issue was the capacity being build that no one used.
It's been 2 years since I could buy a gpu for love or money. It's been one for ram.
My old rig is worth ~two~ three[0] times what I paid for it inflation adjusted.
[0] fuck me ram is rising again.
We just don't know the true value of tokens until those subsidies stop, and the subsidies don't stop because the respective companies don't want to know, at least until everyone cashed out. The valuations of AI companies are very difficult to reconcile with current revenue (which doesn't cover cost) and with a sober mind it's hard to imagine where all this additional revenue is going to come from, down the line. And that's not to say the the technology isn't fundamentally transformative and revolutionary. So was the internet, but it still created an investment bubble.
The RAM prices are in yet another pork cycle, which will be exacerbated by the fact that many previous pork cycles left RAM manufacturers with excess capacity and rock-bottom prices, so they're much more careful investing and more assertive negotiating contracts now.
These models become obsolete in months. Until the labs prove that they can produce a model that has a positive return on investment in it, it’s insane to think that the labs are in a long term good position.
Semianalysis estimates that subscriptions make up 40% of tokens generated for only the 10% of the revenue:
https://newsletter.semianalysis.com/p/anthropic-subscription...
I'm also not saying that this subsidy is irrational, it does buy mindshare, but it still distorts the true demand for tokens. I wouldn't be using anywhere near as many tokens if the price was as high as the API costs, if anything I would switch to a cheaper model, and likely not an Anthropic one.
If you're gonna point to gross margins on the API, those 85% really are meaningless in isolation. They can charge top dollar because they have one of the top models, but only because they keep spending top dollar on training it, as does OpenAI. It's a game of chicken where whoever stops heavy training investments first is likely the one to lose market share to whoever keeps spending. So the play is to try and keep the investment capital flowing and be the last one standing, which only works if you can keep selling a fantasy of future revenue, not the dire reality of current revenue.
Further more, their compute commitments are over 10-20 years. Their average quarterly compute commitment is only about $11B - $15B which they have already surpassed with their revenue.
At $230b/year, they’d be at around $57b in quarterly revenue against compute of around $16b. This is not bleeding cash.
Why do you say they're running out of cash? Can we see your math?
Let's say it's 15 years. That's $16B/quarter in commitment. They had $17B in revenue in Q3 2026.
If anything, they're far too conservative.
most recently (on both a quarterly and annual basis), it was negative
Their revenue pays for the compute.
They can't just turn off all other expenses and just pay for compute.
If VC money gets tight, they would have to issue debt, that pay interests. With a current negative EBITDA, their bond will be junk at IPO, paying higher interests and having a lot less market than AAA debt (e.g. Google has access to that market). Both things combined translate to "running out of cash".
Their S1 is public. We have AI. Are people this lazy? Throw it into ChatGPT and ask it to confirm the numbers.
https://www.sec.gov/Archives/edgar/data/2130357/000162828026...
Income at page F-7 (about 350 overall):
(units in Kilo-millions)
Revenue in first 2026 9-months 33.2
Compute and infra expenses 18.8
Personel 4.8
Sales 5.5
Administrative expenses 5.3
Operating earnings (loss) (1.2)
Q-3 only, this is calculated substracting first half from 9 first months:
Revenue in Q3 16.9
Compute and infra expenses 9.9
Personel 1.7
Sales 2.3
Administrative expenses 3.6
Operating earnings (loss (0.6)
The problem with AI is that it's still dependent on the skill of the prompter.
Ah so when people don’t have the same outlook as you we’ve gone from being lazy to now it’s a skill issue.
It's not exactly laziness. It's that most people don't understand basic financial concepts like fixed versus operating costs.
Companies like Anthropic and OpenAI have invested huge sums in building/training models. Those are the headline numbers everyone talks about. What they don't understand is that the gross profit margins on AI model inference are very high.
The question is whether these companies ever get to the point where the profit from inference justifies (recoups) the investment made in training. If they have to keep investing vast sums in training new models, and/or competition causes the margins on inference to collapse, they're in trouble. But if training becomes less costly and inference margins don't compress too much, it's a totally different story.
Can we be sure of this? The second they tried to hike the prices to be profitable, they lost a lot of clients, or a lot of clients lowered they token expenditure. In any market margins tend to go to zero. I am using frontier models, but only because they are dirty cheap if not free. If they raise the price, I just switch to a cheaper model, start exploring the option of renting a GPU-VPS or buying my own card. Or reducing my AI usage, that currently includes stupid useless things like "translate this cooking recipe from french... for free".
> But if training becomes less costly and inference margins don't compress too much, it's a totally different story.
They are being priced as the "ifs" are certainties, and not only a possibility. If anything goes a millimetre below perfect for them, their finances collapse. High margins attract competitors, and unless you have a huge moat (and their only moat is expensive training), your high margins vanish.
1. Model Build/Training costs are escalating if anything
2. They are in an arms race with competitors and can't afford to stop building new models - if they lose the perception of leading, they lose lots of customers overnight.
3. Inference margin is under attack from more specialised models and open weights models.*
4. They have no moat given how simple the interface to the products is - a bunch of text files and text prompts, and little prospect of building a moat.
So those all run counter the possibility of coasting on previous research. To me the most promising avenue for making money for these companies is to replace Google, not AGI or taking over the world's knowledge work. That would mean they are overvalued. Even coasting on previous research, how long would it take to pay off the money they have spent?
* https://strandsagents.com/blog/amazon-saves-millions-replaci...
> Companies like Anthropic and OpenAI have invested huge sums in building/training models. Those are the headline numbers everyone talks about. What they don't understand is that the gross profit margins on AI model inference are very high.
Only when you ignore that the models they provide inference for have an extremely high cost and pretend they don't depreciate. They keep having to train new models. I don't think they recovered the cost for a model when a new one releases.
It is kind of cute to say other people "don't understand basic financial concepts like fixed versus operating costs" when you seem to not understand basic financial concepts such as depreciation and amortization.
Also, and this is important, we don't really know their true compute cost. We don't know if their compute with the likes of Google and Amazon were initially subsidized, and for how long it might be. In Q2 they also got to not pay SpaceX for compute for a while.
Gross margin is revenue minus cost of revenue (COGS). COGS is basically the compute that serves inference (rented capacity, or depreciation on owned GPUs used for inference, plus expenses like support, etc.).
Model training is normally treated as R&D. Under US GAAP, R&D is expensed as it occurs and is below gross profit as an operating expense. It is not COGS.
You mention "depreciation" but depreciation applies to physical assets like GPUs. The GPU depreciation that belongs in gross margin is already in COGS. A trained model would be an intangible asset, so that is amortization, not depreciation. Even that only applies if the training cost was capitalized, and under GAAP it generally wouldn't be. The full cost is expensed when it's incurred. Nobody is "pretending models don't depreciate." Their cost is counted all at once but not in gross margin because that's simply not what gross margin measures.
Obviously, models lose value fast in an economic sense. They get replaced within a year or two and competition limits what they can earn. But that's an argument about whether the R&D expense is sustainable. It doesn't mean that gross margins are being misstated.
"Inference has high gross margins" and "training is extremely expensive" can both be true, which oddly seems to be something many people have trouble comprehending. But this is the crux of the questions that will determine whether these businesses are viable long-term:
1. Will training costs fall over time, or at least grow more slowly than revenue?
2. Does the gross profit a model earns from inference before it's replaced exceed what it cost to train?
3. Will inference margins hold up as competition and open-weight models push prices down?
You have a legitimate point about compute costs. If the deals with Google, Amazon or anyone else are priced below market, that directly inflates gross margin, because it lowers COGS. But we don't know those terms, so that's just speculation, and it's still a different argument from depreciation.
To reiterate what I wrote:
> The question is whether these companies ever get to the point where the profit from inference justifies (recoups) the investment made in training. If they have to keep investing vast sums in training new models, and/or competition causes the margins on inference to collapse, they're in trouble. But if training becomes less costly and inference margins don't compress too much, it's a totally different story.
They are when they say that inference is incredibly profitable. Its incredible profits exist only when you pretend that models have 0 cost to be created (since they are a digital asset that can't depreciate), and R&D is a imaginary thing that exists only as a fixed cost of doing business that doesn't need to be recovered.
Anthropic released Opus 4.6 in February 2026 and deprecated it in June 2026. Was it anywhere close to recover the "R&D" costs during those few months?
More importantly, how long a model needs to be current until it recovers "R&D"? Their economic viability hangs on this, and the compute costs both to train and servr models.
A lot of people like to vaguely gesture at the possibility of compute costs coming down, but it is in fact coming up with the rising cost of hardware (and possibly with sweetheart compute deals being replaced by contacts where providers actually have to make money).
Expensing the cost of training a model isn't treating it as free. The whole cost is charged to the income statement the year it's incurred, which is actually faster than depreciation would ever recognize it. Calling something R&D does not make the cost "imaginary". It means the cost is counted. It's just not in gross margin.
We are actually asking the same basic question (can the investment made ever be recouped?). What you're struggling with is the accounting concepts and terminology. The costs of creating the models are a different part of the income statement than the costs of serving inference.
Again, "inference has high gross margins" and "training is extremely expensive" can both be true at the same time.
In absolutely none of my comments did I state that R&D never needs to be recovered. In fact, I stated repeatedly that the entire crux of whether these businesses are viable long-term comes down to whether the costs of creating the models is eventually recouped.
> Anthropic released Opus 4.6 in February 2026 and deprecated it in June 2026.
First off, Opus 4.6 is still available, even though Anthropic recommends migrating to 5.5. Google Cloud says it will be on offer until February 2027. So revenue for this model didn't drop to $0 overnight.
https://platform.claude.com/docs/en/models/opus-4-6/overview
More importantly, your comment implies that models cease to have value when the next version is released. This completely ignores the reality of model training, which is that the subsequent models almost certainly built on the research, data and techniques of 4.6. And 4.6 was built on the research, data and techniques of the models that came before it.
So the whole notion that point versions of these models are standalone investments that cost billions to train and then become worthless the minute a new point version is released is incredibly naive. It's just not how model training works.
> A lot of people like to vaguely gesture at the possibility of compute costs coming down, but it is in fact coming up with the rising cost of hardware (and possibly with sweetheart compute deals being replaced by contacts where providers actually have to make money).
You have absolutely no evidence of "sweetheart compute deals" so this is pure speculation.
Furthermore, there's a flaw in your logic. It's entirely possible for chips and memory to get more expensive while the cost per token served falls due to more efficient hardware and better inference software. What matters here is the cost per unit of output, not the price of a GPU.
Bottom line: inference has high gross margins and these businesses have had to invest massive sums. If inference profits grow sufficiently faster than costs, or costs decrease sufficiently over time, these businesses could be viable and perhaps incredibly successful. If that doesn't happen, their viability is questionable.
Lots of people have an opinion about this but as this discussion proves, lots of people also don't understand the basic accounting required to make a cogent argument one way or the other.
The problem I have with this is that the claim that "inference has high gross margins" is paraded around as something that automatically makes all this make sense.
My concern here is something that you described with more nuance, the cost of training cannot just set aside as if it doesn't need to be recovered. R&D makes sense if we consider that research and development eventually gets to a model that doesn't need to be trained anymore, and can be served at relatively low cost indefinitely. Otherwise the constant need to train new models needs to be taken into account; you can only keep revenue by training new ones.
> You have absolutely no evidence of "sweetheart compute deals" so this is pure speculation.
I agree that this is speculation. But it is based on actual sweetheart deals taking place (such as SpaceX not charging for compute during a period in Q2) and on Anthropic being very unclear if their compute cost is based on market rates or subsidized by providers who are also investors in the company.
> Furthermore, there's a flaw in your logic. It's entirely possible for chips and memory to get more expensive while the cost per token served falls due to more efficient hardware and better inference software. What matters here is the cost per unit of output, not the price of a GPU.
I agree that my statement does not take into consideration hypothetical improvements in efficiency, but I dispute that the increased cost of hardware is not a menace to their viability.
You may have a major breakthrough that halves the compute use for a task. It is meaningless if the hardware is now 3 times more expensive.
Except that this isn't the argument I'm making, and I don't think most of the people who are pointing it out are claiming that inference being sold at high gross margins means that you don't have to consider the massive investments that have been made.
The more common narrative is that these companies are losing massive amounts of money and will never be able to make a profit. And that's where looking at the investment in training versus the economics of inference is useful because, once again, inference actually does have high gross margins.
> R&D makes sense if we consider that research and development eventually gets to a model that doesn't need to be trained anymore, and can be served at relatively low cost indefinitely.
This just isn't true. If revenue grows at a sufficient enough rate versus training costs, and/or training costs decrease substantially enough, the numbers could work out over time.
People who are claiming that it's mathematically impossible are going on vibes. And just to be clear: I'm not claiming that it working out over time is guaranteed, or the likely outcome. But it is possible and the mechanics by which it would work out (if that comes to pass) is pretty simple.
> I agree that this is speculation. But it is based on actual sweetheart deals taking place (such as SpaceX not charging for compute during a period in Q2) and on Anthropic being very unclear if their compute cost is based on market rates or subsidized by providers who are also investors in the company.
Regarding SpaceX: my understanding is that as it was ramping up its capacity, which was not optimized and capable of being delivered in full at the time, it provided concessions, including discounts and deferred billing, to large customers. What SpaceX got from this in return was a massive backlog (over $14 billion worth) of non-cancellable contracts.
The concerns about Anthropic investors are legitimate, but again, you can only speculate. If your entire thesis ignores plausible scenarios that aren't so negative (like the possibility that compute is being sold at or near market rates), you have to be honest: the thesis is biased to support the argument you want to make.
If I was going to invest in any of these companies, I would want to know more about the terms of these deals but I don't think anyone who hasn't actually seen the agreements should make confident arguments based on what they think must be in them.
As a side note: if you really want to be exact, we'd need to know how compute purchased at below cost affects the income statement. If the below-market pricing was for compute related to training, it would not be the same on the income statement as compute related to inference. Whereas below-market pricing for inference compute would hit COGS and improve gross margins.
> You may have a major breakthrough that halves the compute use for a task. It is meaningless if the hardware is now 3 times more expensive.
What matters is the price per unit of compute. "The hardware is more expensive" only threatens viability if the price per unit of compute is rising, not just the price per chip.
You're throwing around random numbers to make a point but there is real data on this subject. For example, Epoch AI published a report a couple of months ago that found the cost of achieving a given level of AI performance has fallen about 47% per quarter since 2023 even as hardware and memory prices rose by tens of percent. On one of its science benchmarks the cost per question at a fixed score fell 725x in less than 18 months.
It did also find that large training runs have grown by 2-3x per year, which is the crux of everything, once again: training costs versus inference costs.
Isn’t it plausible that this is conceptually wrong? What if a model’s revenue-generating potential only lasts for something like a year due to, loosely speaking, competition? Doesn’t this make model development more akin to, say, the costs of annual agricultural planting/harvesting/delivery cycles, or a tv network’s annual (back in the day at least…) production cycle for its main content?
Under US GAAP, production costs are capitalized when they're incurred and then amortized into cost of revenue over the period when the content actually earns revenue. Because content usually earns more revenue early and then tapers off, the amortization is usually front-loaded.
The one problem with this is that not all training actually goes into what's sold to customers. There are experiments, failed runs, various studies, etc. So you would need to come up with a way to cost only the final production run (and perhaps post-training) to come up with a number akin to content cost.
ASC 730 basically favors expensing anything uncertain, so unless you have a really good, defensible way of breaking out these costs and separating them, doing it this way would go against the current accounting standards.
the gross margins on growing potatoes is very high, if you exclude all the other costs needed to get the potatoes to market.
Think about it, you take some potatoes you had left over from last yeah, dump them in the ground and wait 9 months. Boom pure profit.
This of course excludes the cost of the land, machinery needed to plant, till and harvest, storage, pesticide, fungicide and herbicide.
Then the transport, cleaning and packaging.
If you arbitrarily cut out all the bits of the business needed to make revenue, everything is a profit.
A company selling potatoes would not include land in COGS. Land is a a fixed asset. Even if a farm rented/leased the land, the payments would fall under OpEx, not COGS.
Seeds, pesticide/fungicide/herbicide, direct labor, water and electricity for irrigation, etc. would all be COGS.
Would it make sense to spend $1 billion on land to grow $1 million worth of potatoes every year? Probably not. But that doesn't mean you couldn't sell potatoes on your $1 billion land at a positive gross margin if revenue exceeded COGS.
This is the AI debate that is impossible to have when people refuse to educate themselves about basic financial accounting concepts.
No man. It’s that, like you’ve explained further in your comment, they haven’t become net profitable yet. Being profitable in 1-2 quarters is a good trend but it doesn’t matter if you are still behind on mountains of debt or investors wanting their money back and your market is one where people talk about shit that happened 4 months ago as obsolete.
There is no proof that the current profit they are making on inference is durable.
Their S1 was taken down. As of today (https://www.reuters.com/world/china/global-markets-view-euro...) This is the second time they changed it.
Apparently they file it with Claude.
IMHO profitability will be worse though. They will contain to take on more and more investments - training, R&D, and 'sales and marketing' for enterprise share. They will face more competition than ever before. The range and types of models will continue to diversify, with real-time models for robotics and computer use paving the way for new entrants.
The investor confidence to back this doubling down will be spurred by rapid progress, panic deregulation while old-guard corporate democrats still hold power in the US, and much stricter sanctions against Chinese competitors, their models, and their eventual domestic hardware. The tech world will look a lot more factional by that point, with pseudo mergers & acquisitions amongst the big companies beginning, unabated by a toothless FTC. Anthropic will probably get closer to Amazon, and we could see big moves amongst Meta, Microsoft, and/or Google too.
Ultimately, Anthropic and OpenAI will maintain their market leading positions through at least 2028/2029 but only after a protracted period of significant equity dilution due to the expensive acquisitions and large outside investments which will be necessary in these market conditions.
Chinese hardware and models will look like a serious threat by the end of 2027 but will be largely locked out of US/Western markets. At the point the equation will have changed drastically - credit issues in the West led by mortgage security collapses will shape the investment environment and 2028 election, and rapid developments in physical technology and manufacturing will take the spotlight away from intelligence itself.
Looking at their 2027 compute commitments, it looks like around $72B total.
So I don't understand where this "burning cash" claim comes from by a few people here.
People lack the accounting know-how to talk about it correctly. But compute for inference is just not the main expense right now.
If they take on more funding for training and/or build out, then they have to share the profits with more people. The growing revenue is shared by more people. As more options vest, they have to share with more people. If they need to acquire external companies for particular capabilities, profit is shared. If they have to pay hand over fist for hyperscaler compute in 2027 - more profit shared.
So as their revenue grows rapidly, the fraction of that revenue which they've promised to other people is growing too - case in point, they operated at an $80B loss on paper this year because the unrealized future-profits promised to early investors is so valuable relative to current revenues. If they have to make more such deals to continue investment (e.g a low IPO price which require shelling even more shares to stay competitive), then they will continue to create vast value but less of that value will be captured by the current investors (the 'burning cash' may go to compute providers, outside investors, as discounts to enterprise and consumers, etc instead of to Anthropic).
Growing 70b revenue to 700b revenue is, ah, a smidge tougher. Even getting $70b to $230b would be an immense success, methinks. The initial hypecycle seems to have cooled down slightly, and people are no longer ranked by how many tokens they can burn through with no real deliverables expected. There's even reports that companies are starting to tighten their belts when it comes to token spend; a catastrophe for fat US labs that need those big-wallet companies to spend API prices which pays for everything. Subscriptions might make them some money but it's not where the real profit is.
Which isn't to say that the revenue won't grow, or even that it'll contract, but I certainly don't see Anth growing to eye-watering $700b next year, especially as long as open-weights models keep getting released and a BigCo can't get Opus deployed to their own hardware sitting in their basement with no chance of the secret sauce leaking out to anybody, including Anthropic.
Anthropic operates at a loss. They lose money.
That is what they keep having to raise money and are desperate to IPO as soon as possible.
> Why do you say they're running out of cash? Can we see your math?
My brother in Christ, you can't tell the difference between revenue and profit, and think that just because their revenue (not profits) grew 10x that they can grow now a "measly" 3x.
They might have enough money (ot whatever they consider to be equivalent) in hand to keep being unprofitable for a bit, but they alao have something in between half a trillion and a trillion of contracts for compute in the next few years.
Edit: And Anthropic was never a specific point of discussion until you brought it up.
You can use Newton for most stuff and get pretty good answers. Micro is questionable in all kinds of ways. Assumptions about actors, assumptions about responses, assumptions about how utility maps onto monetary value. It's a huge mess really.
The money is there but not at the risk levels these companies have pushed things to. They're spending billions like it's candy, and treating millions like change they found in their couch cushions.
Financial systems are usually in the hands of people that have some risk management. When these labs say things like this will change everything, there's no way to know how to evaluate any of this.