Anthropic Sets IPO Price Range at $70–$78, Valuing Claude Maker at $1.3 Trillion

What Actually Happened

On Tuesday, Anthropic filed a price range for its initial public offering: shares at $70 to $78 each, 825 million of them, implying a valuation of about $1.3 trillion, according to Forbes and Reuters. At the top of the range the raise lands at roughly $64.4 billion, which Reuters and CNBC described as the largest AI public offering on record. Anthropic priced where it did because the market forced the question: after months of private-round chatter about higher numbers, the company is meeting Wall Street's demand for a real clearing price, not a headline.

The honest read is that the final range came in below the $1.5 trillion figure that some investors had circulated in private, per Bloomberg and Reuters — a delta of roughly $200 billion. That gap is the story. A year ago, AI valuations ratcheted up by name recognition; this filing shows a frontier lab trading at a discount to its own hype rather than riding it.

The Number Behind the Number

Reuters put the top-of-range value at $64.4 billion in proceeds, which would make the Claude maker one of the most valuable companies on Earth at listing, ahead of the traditional tech bellwethers that built their empires over decades. The share count — 825 million — is itself a tell: this is a broad distribution, not a controlled float designed to manufactured scarcity.

Anthropic's own framing in the filing struck a split tone. NBC News reported the company's stated purpose was to "bring forth an age of miracles," language lifted from the secular-salvation register of tech IPOs past. In the same document, however, it warned of existential risk to humanity, per NBC News. The tension between growth prophecy and risk disclosure is not cosmetic: it is the document public-market rules require, and it is rarely that explicit.

What Builders Should Take From It

• The cost of capital is now visible. Frontier AI was priced in private; now it has a ticker. Every contract you sign against a model, every Token-use assumption on your roadmap, can be checked against a public market value rather than a rumor.
• The valuation delta matters. A $200 billion gap between floated and filed numbers says investors did not swallow the top estimate wholesale. For builders, that is a reason to price your own AI margins against a range, not a point.
• Disclosure becomes a standard. Once a frontier lab reports revenue, compute cost, and customer concentration publicly, the assumptions behind model pricing will be testable rather than anecdotal. Expect pricing tables to sharpen within a quarter or two.
• The existential-risk footnote is new. Public-market risk factors that name humanity-scale harm are unusual; they give procurement teams a document they can point to when they ask their counsel what they signed.
• Competition got a price tag. OpenAI, Google DeepMind, Meta, Mistral now sit a measured distance from a trillion-dollar public neighbor. Stack choices made this quarter will be re-run against that neighbor's margin structure.

Closer

The trillion-dollar listing is a milestone for the industry, but the more useful number for builders is the gap: $1.5 trillion expected, $1.3 trillion filed. That gap is what discipline looks like. The rest of us get to price our software against it.

Developer312 covers the AI business signals builders actually need to act on. Get the weekday briefing at developer312.com.