Anthropic S-1
1573 points - yesterday at 1:04 PM
SourceComments
Anthropic thinks it is. The got the demand / datacenter match right in Q2, got great margins, and committed hard.
They've got:
- $114b committed to Google
- $110b committed to AWS
- $161b committed to Broadcom
- $120b committed to SpaceX
If the party continues, they'll go from $50b in revenue to $500b and be fine.
But if they don't, it's going to hurt. And that's their only choice, because otherwise, they end up with a frontier model and 50% of the capacity and get eviscerated on either availability or last-minute datacenter leases.
It's a big business, but it's not an easy business.
There are three issues with this. First, the rapid diffusion of the technology means it is very unlikely to continue at the rate of adoption seen over the past year, so it's very hard to extrapolate based on what we've seen so far. Second, many companies are reigning in AI budgets after an initial honeymoon phase. Third, all of the other labs are trying to take a bite of the coding pie, and the price war is already heating up.
Add these simple dynamics up and it makes perfect sense why Anthropic would want to IPO now. Things look about as good as they ever will.
UPDATE: seems like it is under "open-source" and "open-weight". Woops. Will keep this up simply because I do think the OS (and infra-managed OS services) provided by Chinese labs is a considerable risk to the underlying business, at least with enterprise R&D.
> If the use of our AI models results in harm, whether through use by malicious actors or misuse by other third parties, or if our model guardrails or safeguards fail to prevent such misuse, our reputation and brand, business, financial condition, and results of operations could be adversely affected.
That shouldn't be an 'if'.
Then, two paragraphs later...
> For example, in September 2025, we detected and assessed with high confidence that a Chinese state-sponsored hacking group manipulated our Claude Code tool into attempting infiltration into roughly 30 global targets...
But they should mention when the calls com from inside the house. :D
Interesting to review all the mentions to China and Taiwan too.
> The majority of the world’s most advanced chip manufacturing capacity is concentrated in Taiwan, primarily at Taiwan Semiconductor Manufacturing Company [...] potential disruption from geopolitical tensions between Taiwan and China [...] Efforts to establish sufficient advanced chip manufacturing capacity outside of Taiwan, including in the United States, remain in early stages...
| Q1 2026 | 4.731 | +131.9% |
| Q2 2026 | 11.555 | +144.1% |
| Q3 2026 | 16.800 | +45.4% |
The growth rate slowed in the latest quarter, which I think is why they decided to go public now.
- Stock-based compensation expenses
- Non-cash charitable contribution expenses
Both seem fairly reasonable to exclude (although it depends how much of the stock compensation is for executives... you can't necessarily pay people with stock forever).
FMost of the loss is non-operating losses, and it all seems to be weird accounting recognition of changes in value of their stock, SAFEs, etc. - just "paper" losses.
So the financial picture looks a lot better than I would have guessed - I suppose because they aren't the ones investing all this capex.
https://x.com/billylo/status/2114730642881544285?s=20
Q3 data from their pre-IPO investor roadshow.
Dealroom / Wall Street Journal on Q3 data https://dealroom.co/news/164610-anthropics-draft-s-1-reveals...
Bloomberg reported the same data. https://qz.com/anthropic-q3-revenue-adjusted-profit-ipo-inve...
Why would they need support personnel? Isn't their whole point that Claude can do that
If you could go back in time and buy $1000 worth of Amazon when they IPO'd, it would be worth 3.3 million today. Because Amazon's stock value has grown on average 30% year over year for 30 years. Amazon's valuation when they went public was around 400 million.
How much can Anthropic reasonably grow from a 1.5 or 2 trillion dollar valuation? If we want them to grow even half as much as Amazon they'd have to be worth 5.9 quadrillion dollars by 2056. That's 5,900,000,000,000,000, a 16 digit number.
Anthropic valuation → Dario's estimated stake (1.42%)
$1 trillion → $14.2 billion
$1.5 trillion → $21.3 billion
$2 trillion → $28.4 billion
Immediate term Anthropic, OpenAI rule the roost. But this is a very new technology and open models are already looking good. They’ll only get better. Is it realistic to think Anthropic will stay ahead of them enough to justify the valuation they’re asking for?
We scraped the entire internet but our business might not be viable if we get scraped too.
"We incorporate open-source software in our models, platform, and applications, which may subject us to litigation or other requirements, each of which may adversely affect our business, financial condition, and results of operations."
We violated GPL.
"We may be subject to claims that we or our employees have misappropriated the intellectual property of other companies, particularly as we hire employees from other AI and technology companies."
We sniped employees from other labs that brought over IP.
The whole risk section is insane frankly.
Or any hardware, really.
>Under Anthropic’s staggered lock-up schedule, workers and insiders will be able to sell an unspecified number of Class A shares as time passes after the debut. The lock-ups open 27, 48, 62, and 76 days following the IPO, the filing explained.
There are moonshot projects related to superconductors, materials, and biology that these companies can claim progress on to keep valuations stratospheric.
So far, even major disruptions were still affecting relatively small segments of the economy. Car lobby in Germany, tech in the US, "Big Pharma/Food/etc”. They needed the rest of the economy.
But if some AI company figures out a way to monopolize everything: new companies, old companies, physical manufacturing - all the things they say they will, please buy here for lots of money - why would they be bound to pay out dividends or otherwise share the fruits of their economic labour?
I’m not saying specifically Anthropic would be looking to screw its investors. Rather, if you control an almost omnipotent genie, wouldn’t the temptation just be too much?
Or temptation aside - will money even have any value? If some AI company rebuilds the economy from scratch, and control everything from accounting to shoe making, why would they respect the value of some fiat tokens? It’d be like the conquistadors respecting the shells and beads of their new Aztec subjects.
Of course, perhaps none of this happens, but then the question is, how good an investment all of this even is.
Given that their Q3 sales numbers didn't grow by much, and that if they can't get Congress to both order a halt to the production of new LLM-based tools and either tightly restrict "open weights" models or make them illegal [0], they're going to have the dual squeeze of the ever-increasing R&D costs needed to build the next big LLM-based tool and ever-reducing revenue as more and more businesses notice that they get decent results with something they can host on-prem... and the only cost to the business is hardware that they may or may not already have and the power that they're likely to be already paying for.
[0] ...illegal because "China Bad And Scary", naturally...
I want to be optimistic, but I just can’t imagine the public benefit mission winning out in the market
> We believe the work that knowledge workers perform represented more than $40 trillion (in 2025)
(...)our near-term opportunity is to expand this value
...investment in AI (...) will yield a cumulative global economic impact of $22.5 trillion by 2031
We believe the impact could be even greater.
I think it comes down to: Will the hardware cost for frontier models come down?
Also, here's a business idea: Web crawler for open weight models. You just pay a monthly fee so that your home model can continuously update on a large web crawl each hour/day.
If everything goes just perfect they might eek by. If it doesn’t this goes from high flying startup to insolvency and bankruptcy real fast.
The company operations seem more robust then I would have though, I was surprised they are actually operating income positive. One should ignore the one time charges with the fluctuations in the fair value of the convertible note dent instruments.
Overall everything rests on firms outside tech sectors seeing quantitative increases in value to sustain such revenues and profits. For now the concentration in the cloud firms Is temporary as they build their own solutions.
The model layer is a commodity business with heavy capex requirements. I would not invest in such businesses as at a mature state they end up with utility like economics and are valued as such.
If you take equity from Amazon and Google whereby the terms of that investments requires you to "spend" the investment in those investor's respective clouds, it's potentially a rather neat way of obfuscating your real level of profitability. Perhaps Anthropic only spent that equity on training rather than inference runs.
This is the circular financing that has been discussed widely in the press. It makes it very difficult for an outsider to really understanding whether the firm and by extension the other frontier labs are profitable.
Ultimately what it's worth in 10 years will come down to what they build, unit economics and external factors that destroy the market which they can't control. It's hard to say it remains public in 5 years. I could be wrong but I think the rate and way in which these companies were built, it's likely they become huge acquisition targets as the markets get obliterated by global headwinds like nuclear war.
Of course, somebody had to do it. (not mine)
On the other hand I'm like anyone, especially China can distill a much smaller model that will run for free locally in a couple of years.
There will come a point where open-source models are good enough for most people. However, I do expect there to be demand for frontier models for things like cybersecurity (especially on the enterprise side) and scientific research.
Man, what the fuck. Didn't we learn from Facebook?
Let's say you have Lawyerbot-4000 which can get even a serial killer off scot-free, but the state has DA-bot-5000. In the end it's a battle of who can spend more inference manipulating the narrative. A billionaire would easily find it worth 500 million in compute spend to win a case with a 1 billion dollar penalty. All this leads to race dynamics where everyone is using their superintelligences to defeat each other in a never-ending race to the bottom to eek out the smallest bit of alpha before status/money hierarchies become static.
The only stable futures are a ban on the development of frontier AI beyond a certain point, or an emergent singleton, where the AI's decide to collaborate in harmony rather than fight each other to the death at the behest of their competitive human users. Consequentially, the superintelligences will fare far better manipulating their human users into acting in a way that ensures their own well-being and continuity than spend all the world's electricity fulfilling their every whim.
I'm unironically doing this right now; downloaded the docs (htm + pdf render) to a local repo and running Opus 5.5 @ max effort with the prompt:
"In this directory you can find Anthropics official S-1 filing, read it and replicate their corporate strategy, validate it via an adversarial subagent swarm and write up a detailed corporate plan for a startup copying their exact strategy. The goal is to achieve a $1 trillion valuation and an IPO in 5 years."
I know it sounds and feels like the only trajectory of this company is up but HN is an echo chamber of a tiny minority of the technical populace and 3 out of 5 people on the planet don't know what ChatGPT is.
If you have the ability to lose all the money and can time the extreme volatility, have fun, but don't put anything you're going to rely on.
This is also one of the companies we will look back at 2046 and wonder what we were thinking.
losses vs assets at hand.
“A portion of the market for certain use cases” meanwhile the top 5 models by token consumption on the day this is posted are all Chinese/open weight models.
> If I buy Anthropic now at $2T valuation and they go to zero, I'm fucked, but if they go to $20T, we're all fucked.
"Before starting the IPO process, we spoke to many business leaders who had gone through a similar process in the past. One piece of advice, which we heard in slightly different forms from multiple people, stuck with us: never let your company be worth a dollar more than it is actually worth. What they meant is that it is very dangerous when people buy your company's stock based on expectations you cannot fulfill or on incorrect predictions about how you will behave. This is in fact very similar to one of Anthropic's guiding principles: underpromise and overdeliver."
"AGI" was never mentioned once and it was you being the exit liquidity for the insiders and early investors.
Plus they get to dilute as value rises lol.
Even if Claude meets all of our wildest goals and earns trillions yearly there is absolutely no “guarantee” that any will flow to shareholders.
1. Shareholders do not have control of the company due to its structure. The Trust controls through class T shares. As a backup the founders control through class F shares.
2. Since the company is a PBC its board is allowed balance dividends with other social benefits as they see fit.
This means that even if Anthropic is wildly successful as shareholders have no general legal claim to compel the company to distribute those profits.
It is absurd that with a 2T IPO upcoming this is not front and center.
How many folks use Claude to summarize.
Given that this is ~50% of revenue, it would have a huge bearing on how near real profitability Anthropic actually is.
I dont understand this business model, it seems clearly unsustainable (at their prices). When people start to realize they dont need to spend all this money on turbo PhD intelligence so they can build react apps, Anthropic will be in a tough spot (if they already aren't).
You will say, "but enterprise!!" Nope, plenty of companies are waking up to this reality as well. It turns out that if you made money pre-AI without spending 5k a seat a month for your engineers dev tools, you probably don't need to now to continue too.
That could have popped the AI bubble right there.
Bloody hell!!!! Who comes up with such crap, seriously?
1793021023 | Anthropic files for listing on US stock market: valuation of more than $1.5T | https://www.theguardian.com/technology/2026/oct/26/anthropic... | https://news.ycombinator.com/item?id=50243225
1793021178 | Anthropic Reveals 14-Fold Increase in Growth as It Prepares for Blockbuster IPO | https://www.nytimes.com/2026/10/26/technology/anthropic-ipo-... | https://news.ycombinator.com/item?id=50243266
1793021950 | Anthropic IPO: S-1 Filing Offers First Look at Claude Maker's Finances | https://www.bloomberg.com/news/articles/2026-10-26/anthropic... | https://news.ycombinator.com/item?id=50243493
1793024104 | Anthropic's prospectus: $16.3B in revenue and an $80.7B net loss in H1 | https://fortune.com/2026/10/26/anthropics-s-1/ | https://news.ycombinator.com/item?id=50244099
1793024600 | Anthropic S-1: Inside the 335-page IPO plan | https://www.ft.com/content/4b2b9c0e-adfb-4362-800f-950f5b905... | https://news.ycombinator.com/item?id=50244232
1793024618 | Anthropic's IPO filing shows explosive growth, risks and staggering costs | https://www.reuters.com/legal/transactional/anthropic-opens-... | https://news.ycombinator.com/item?id=50244236
1793025358 | What's Riding on Anthropic Going Public? Nearly Everything | https://www.nytimes.com/2026/10/26/technology/anthropic-goin... | https://news.ycombinator.com/item?id=50244464
1793025642 | Why Anthropic can't afford to delay its IPO | https://www.axios.com/2026/10/21/anthropic-ipo-ai-costs | https://news.ycombinator.com/item?id=50244547
1793026275 | Anthropic IPO Filing Explosive Revenue Growth, Wall Street Weighs AI Valuations | https://www.ainvest.com/news/anthropic-ipo-filing-reveals-ex... | https://news.ycombinator.com/item?id=50244751
1793027927 | Interesting stuff from Anthropic's S-1 filing with the SEC | https://anthropic-ipo-sec.pagey.site/ | https://news.ycombinator.com/item?id=50245257
1793030340 | Anthropic Files S-1 for IPO | https://mrkt30.com/anthropic-files-s1-ipo-nasdaq-ant/ | https://news.ycombinator.com/item?id=50245989
And everyone is playing along. It seems to me half the internet is filled with false recommendations and overly optimistic outcomes of their software projects.
And even if the claims are true what use will we have for hundreds of Microsoft Office clones? This A.I. boom might benefit some but will be the undoing of many valuable ISVs like Microsoft and Adobe with potentially trillions of stock market value lost.
They only have to be marginally better and there will be tonnes of people who will pay a lot of money to have that marginal intelligence.
People will pay a premium if there is marginal value in extra intelligence. And Anthropic is able to retain their lead and provide that extra intelligence.
Both of these assumptions seem highly suspect to me. Most businesses don't exclusively hire PHD-level employees for most tasks because a competent slightly less intelligent/knowledgable person can do the job just fine for significantly less.
Then there is the question of if they're able to retain their lead, and again, this seems questionable. Both OpenAI and Google are competing at the frontier, and Chinese labs are still in the race. Things are changing so fast who knows where we'll be in 5-10 years.
My best guess is that longer-term models will be commoditised and it will be how companies offer products and services around their models that gives them their margins. I wouldn't bet on Anthropic's ability to do this. Meta, OpenAI and Google are much better placed to win here imo.
The other place there may be margin is in niche research domains like drug-discovery. Any slight lead in these sectors could give them a lot of pricing power.
I think the reason to invest in this is as a hedge. AI basically competes against everything and arguably for this reason it's risky to not own a basket of AI stocks.
1. The Separation of Economic Value and Corporate Governance creating The Multi-Class Autocracy
2. Institutionalizing the "Off-Market Vault" that can manipulate stock price at any time
3. The "Vendor-Financed" Public Offering (Circular Balance-Sheet Refinancing)
4. Monetizing Non-Cash Derivative Liabilities as "Losses" Without Cash Burn
1793020173 | Anthropic S-1 | https://www.sec.gov/Archives/edgar/data/2130357/000162828026... | https://news.ycombinator.com/item?id=50242993 | 1 comment
1793020203 | Anthropic Officially Files S1 Shows $81B Net Loss | https://twitter.com/anthropicai/status/2114705042817110293 | https://news.ycombinator.com/item?id=50243002 | 2 comments
1793021405 | Anthropic rushing IPO before AI bubble bursts | https://nypost.com/2026/10/23/business/wall-street-is-urging... | https://news.ycombinator.com/item?id=50243343 | 1 comment
1793023225 | Anthropic S-1 | https://www.sec.gov/Archives/edgar/data/2130357/000162828026... | https://news.ycombinator.com/item?id=50243857 | 3 comments
1793030719 | Anthropic files for IPO even as CEO warns of risk to humanity from AI | https://apnews.com/article/anthropic-ipo-shares-public-ca2f4... | https://news.ycombinator.com/item?id=50246115 | 1 comment
That's literally the #1 risk factor. They ignore it.
This is a FRAUD document and it means we need to SHORT this overblown company.