Updated 19 November 2026 — the Anthropic IPO is priced. Anthropic set its final price at $83 a share on 18 November, above the $70 to $78 range it filed a week earlier, and its shares begin trading on the Nasdaq today under the ticker ANT. This guide explains what happened, what the filing actually says, and what it means if Claude is part of your workflow. It is educational content only, not financial advice.
The company behind Claude just showed the world its real numbers — $16.3 billion of revenue in the first half of 2026, then $16.96 billion in the third quarter alone — and the market answered with the largest IPO on record. Anthropic sold 980 million shares at $83 each, raising about $81.3 billion and valuing the company at roughly $1.5 trillion on a fully diluted basis.
You do not need to buy a single share to be affected by this. If Claude writes your emails, drafts your proposals, or powers a service you sell, the company behind it has just changed shape — from a private lab answering to a handful of investors, to a public company answering to the stock market every three months. That changes the incentives around pricing, the free tier, and the skills clients will pay for. Here is what actually changed, and what did not.
In this guide:
- What an S-1 and an IPO actually are
- The final numbers: $83, 980 million shares, $81.3 billion
- The five numbers from the filing that matter
- What public-company pressure usually does to AI pricing and free tiers
- Why a public Claude means more demand for Claude skills
- The certification angle: why Claude credentials matter more now
- What not to do: scams, FOMO and day-one buying
- A calm watchlist: the dates that actually matter next
- Frequently asked questions
What an S-1 and an IPO actually are
An IPO — an initial public offering — is the first time a private company sells shares to the public. Before it, only venture funds, staff and a few big partners own a piece. After it, anyone with a brokerage account can buy in, and the share price becomes a public scoreboard the company is judged on daily.
The S-1 is the document that makes it legal. It is the registration statement a company files with the US Securities and Exchange Commission, and it is where the company has to tell the truth in writing: audited revenue, losses, debts, who owns what, what could go wrong, and how much its bosses are paid. Anthropic filed its S-1 on 26 October 2026 — 335 pages that contained the first verified figures outsiders had ever seen about the business.
Two more filings followed. A first amendment on 2 November added the third quarter. A second amendment on 10 November set the price range — $70 to $78 a share for 825 million shares — and launched the roadshow, the two-week tour where bankers pitch the deal to big investors and take their orders. A third amendment on 17 November updated the ownership picture. Then, on the evening of 18 November, Anthropic announced the final price: $83, with the deal upsized to 980 million shares.
One thing worth knowing: the filing itself never set the dates. The expectation that pricing would land on 18 November and trading would start on 19 November came from Anthropic’s own IPO materials and press reports, and it held. The SEC declared the registration effective on 18 November, and trading under the ticker ANT begins on the Nasdaq today, 19 November. The offering is expected to close on 20 November.
The final numbers: $83, 980 million shares, $81.3 billion
Demand beat the range. Bloomberg reported the order book ran to roughly seven times the shares on offer, and Altimeter Capital’s Brad Gerstner said on CNBC on 16 November that his firm alone had bid $5 billion. Orders are indications of interest, not firm commitments, but the direction was clear enough that Anthropic and its bankers raised both the price and the size of the deal.
| Term | As filed on 10 November | Final, priced 18 November |
|---|---|---|
| Price per share | $70 to $78 | $83 |
| Shares sold | 825 million | 980 million (956.4 million new, 23.6 million from existing holders) |
| Amount raised | Up to about $64 billion | About $81.3 billion in total; about $79.4 billion to the company |
| Valuation, fully diluted | Roughly $1.26 to $1.4 trillion | Roughly $1.5 trillion |
| First trading | Expected 19 November | Today, 19 November, Nasdaq as ANT |
At $81.3 billion, this is the largest IPO on record, passing the previous record set earlier in 2026, according to Reuters. The underwriters — Goldman Sachs and Morgan Stanley lead the deal — also hold a 30-day option to buy up to 147 million more shares, which would take the total higher still if demand holds.
Around a fifth of the offering was set aside for retail investors, according to CNBC, which is unusually generous for a deal of this size. That does not mean retail buyers got a bargain: $83 was the price for everyone, and what the shares are worth after the first day of open trading is whatever the market decides.
Alongside the public offering, seven of Anthropic’s biggest suppliers — AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC — are buying about $26 billion of shares in a concurrent private placement, at the same $83 price. Those are non-voting shares, but the message is plain: the companies that sell Anthropic its chips also want a stake in it. After the offering, Amazon holds about 14.7% of the company and Google about 13.7%, according to the third amendment filed on 17 November.
The five numbers from the filing that matter
Most of the S-1 is legal boilerplate. Five numbers do the real work.
1. Revenue went from $1.16 billion to $16.29 billion in a year. That is the first half of 2026 against the first half of 2025 — roughly fourteen times bigger. The third quarter alone brought in $16.96 billion. For the first nine months of 2026, revenue reached $33.2 billion, against $2.5 billion a year earlier. This is the number that explains the valuation: investors are paying for a growth curve, not for today’s profits.
2. The business made its first operating profit in the second quarter. Then the third quarter showed a $567 million operating loss. Read together, that says the core business can be profitable, but Anthropic is choosing to spend heavily — mostly on computing power — faster than the revenue arrives. Which brings us to the next number.
3. About $895 billion of compute and infrastructure commitments. Anthropic has promised to spend roughly that much on the data centres and chips it needs to train and run its models. It spent nearly $9 billion on infrastructure in the first half of 2026 alone. This is the other side of the IPO: the money raised is, in large part, money to pay for computing capacity. If you use Claude, this is the number that decides whether the service stays fast and available as demand grows.
4. The $80.7 billion “loss” is mostly a paper charge. The headline loss for the first half of 2026 looks catastrophic until you read what it is made of. Most of it is a non-cash accounting charge for stock-based compensation — shares promised to staff that became far more valuable as Anthropic’s own valuation rose. No cash left the building for that part. It is still a real cost to shareholders, because those shares dilute everyone else, but it is not the same as a company burning $80 billion in cash. This is the single most misunderstood figure in the whole filing.
5. The founders keep 50.1% of the votes. Anthropic’s founders retain just over half the voting power after the listing. In practice that means public shareholders own most of the economics but cannot outvote the founders on the big decisions. For a user, that cuts both ways: it protects the company’s long-term direction from short-term market pressure, and it means ordinary shareholders have little say if that direction changes.
What public-company pressure usually does to AI pricing and free tiers
Here is the honest part. A public company reports earnings every three months, and every quarter the market asks the same question: did revenue grow, and did margins improve? That creates a steady pressure to charge more, to move free users onto paid plans, and to cut the parts of the product that cost a lot and earn nothing.
So far, nothing has changed for Claude users. We checked Anthropic’s pricing page on the morning of the listing, 19 November, and the plans are exactly where they were: Free at $0, Pro at $20 a month ($17 a month if you pay $200 up front for the year), Max at $100 a month with five times Pro’s usage, and Ultra at $200 a month with ten times Pro’s weekly usage. The free tier is still there, and Claude Code is still included in every paid plan. If you want the full breakdown of what each plan actually buys you, our Claude review walks through it with the trade-offs spelled out.
What is worth watching, rather than worrying about today, is the shape of future changes. Three patterns show up again and again when fast-growing tech companies go public: usage limits get tighter on the free tier before prices rise on paid tiers; the most powerful models drift towards the most expensive plans; and features that were once bundled start appearing as add-ons. Anthropic has already moved in that direction at the edges — its newest features, like Claude Sheets, Whiteboards and Claude Data, arrived on paid plans only — and the Fable model now sits behind usage credits on Pro.
None of that is a prediction. It is a pattern, and the sensible response is boring: keep an eye on the pricing page when your plan renews, and do not build a business process that only works on a free tier. If Claude is load-bearing in your work, the $20 Pro plan is the cheapest insurance against a free-tier change, and it is the plan most small businesses should be on anyway for the higher usage limits.
Why a public Claude means more demand for Claude skills
The IPO does something useful for anyone who sells Claude-related work: it makes Claude a boardroom-safe choice. Before the listing, a cautious IT manager could reasonably ask whether a private lab with no public accounts was a safe bet for a company-wide rollout. After it, the numbers are public, the governance is public, and the biggest names in tech are shareholders. That lowers the bar for enterprise adoption, and enterprise adoption is what creates paid work for people who know the tool.
The evidence was already there before the listing. Cognizant announced on 10 November that it had built a Claude Frontier Capability Group of more than 25,000 professionals delivering Claude to enterprise clients, with 44,000 employees trained on Claude and more than 26,000 Claude certifications held across the firm. One company alone holding 26,000 certifications tells you where the hiring is heading: clients do not just want the software, they want people who can prove they know how to use it properly.
For a freelancer or small agency, the practical read is simple. The skills that will be in demand are not “can you prompt an AI” — that is table stakes now. They are the unglamorous ones: setting Claude up inside a real workflow, writing the instructions and guardrails that keep it reliable, connecting it to the tools a business already uses, and showing a client the result in money saved or time recovered. If you can do that, the IPO just made your market bigger.
The certification angle: why Claude credentials matter more now
Certifications are easy to dismiss until a hiring manager has to choose between two strangers. A public listing makes Claude a standard enterprise tool, and standard tools get standard hiring filters — including credentials. The Cognizant numbers above are the clearest signal: when a firm of that size trains 44,000 people and counts certifications in the tens of thousands, it is because their clients ask for certified people.
That does not mean a certificate alone gets you hired. It means it gets you past the first screen, which is where most applications die. If you already use Claude daily for client work, formalising that into a credential is cheap compared with the work it can unlock — and it is a lot cheaper than learning a new tool from scratch. The honest trade-off: a credential proves you passed a test, not that you can deliver for a real client. Pair it with one or two documented examples of work you actually did, and it carries real weight. On its own, it is a line on a CV.
What not to do: scams, FOMO and day-one buying
Big IPOs attract the same three traps every time, and the Anthropic IPO is no exception.
Pre-IPO share scams. If someone messages you offering Anthropic shares before the listing, a guaranteed allocation, or a “private round” for small investors, it is a scam. The offering was sold through Goldman Sachs and Morgan Stanley to institutional and retail channels — not through WhatsApp, Telegram or a stranger’s link. Anthropic’s own IPO page is the only place to check the facts, and no legitimate allocation ever arrives by direct message.
Day-one FOMO. The first day of trading is the worst day to make a calm decision. Prices on debut day are driven by hype, scarcity and headlines, not by the filing you just read. Shares that open far above the IPO price often give some of it back in the following weeks. If you are considering buying, the filing will still be there next week, and the price will be set by the market rather than the opening auction. Nothing in this article is financial advice, and nobody should buy with money they need for rent, bills or a business expense.
Betting on the story instead of the numbers. The story is seductive — the fastest-growing company in the history of the AI boom. The numbers say the same company has committed to spending about $895 billion on computing, is not yet consistently profitable, and is valued at roughly $1.5 trillion before a single day of public trading. Both things are true at once. Anyone who only tells you one of them is selling you something.
A calm watchlist: the dates that actually matter next
If you want to follow this without refreshing a stock ticker, four dates matter.
- Today, 19 November: first trading on the Nasdaq as ANT. The opening price is set by the market, not by Anthropic — watch where it closes, not where it opens.
- 20 November: the offering closes, and the underwriters’ 30-day option on up to 147 million extra shares starts running.
- February to March 2027: Anthropic reports its full-year 2026 results. That is also when the next chunk of the lock-up releases — about 35% of locked-up shares become sellable after the results. More shares available to sell can mean downward pressure on the price, and it is the first real test of whether the growth story survives a public earnings call.
- Around May 2027: day 180 after listing, when most remaining locked-up shares become sellable for ordinary holders. Founders and the largest holders stay locked up longer, in some cases up to a year.
The lock-up matters because it controls supply. On day one, only a fraction of Anthropic’s shares can actually trade — about 10% of locked-up shares were releasable from the start — so early prices reflect a small slice of the company. As each lock-up chunk releases, the real supply appears. If you are a Claude user rather than an investor, the earnings date is the one to note: it is the first time the market will judge the company on results rather than on the roadshow story, and that judgement is what shapes pricing decisions afterwards.
Frequently asked questions
What is Anthropic’s IPO price?
Anthropic priced its IPO at $83 a share on 18 November 2026, above the $70 to $78 range it set on 10 November. It sold 980 million shares and raised about $81.3 billion, making it the largest IPO on record.
When does Anthropic start trading, and under what ticker?
Anthropic’s Class A shares begin trading on the Nasdaq Global Select Market on Thursday 19 November 2026, under the ticker ANT. The offering is expected to close on 20 November.
Has Claude’s pricing changed because of the IPO?
No. As of the morning of the listing, 19 November, Claude’s plans are unchanged: Free $0, Pro $20 a month ($17 a month on annual billing), Max $100 a month and Ultra $200 a month. Going public creates pressure to monetise over time, so it is worth checking the pricing page at renewal, but there has been no change so far. Our Claude review has the full plan comparison.
How much is Anthropic worth after the IPO?
Roughly $1.5 trillion on a fully diluted basis at the $83 IPO price — that counts all shares including options and unvested stock. On shares actually outstanding after the offering, the value is closer to $1 trillion. The fully diluted figure is the one most headlines quote.
Can I still buy Anthropic shares?
Yes, once trading opens on 19 November you can buy ANT through any broker that offers US stocks, at the market price — which may be above or below the $83 IPO price. About a fifth of the offering was reserved for retail investors. This article is educational only and is not financial advice; never invest money you need for living costs or your business.
Sources: Anthropic’s S-1 and amendments on SEC EDGAR; Anthropic’s pricing announcement of 18 November 2026; Claude pricing, checked 19 November 2026; Reuters reporting on the pricing via FXStreet; filing tracker with the full timeline. Figures in this article are as filed or as announced; the market price of ANT after listing will differ from the $83 IPO price. Also useful if you use Claude for coding: our comparison of coding-agent prices ahead of the 1 December rise, and the context behind the compute spending in the AI factory boom.