The Biggest IPO in History Is Weeks Away. Here's What I'm Watching.
Anthropic is heading to the Nasdaq with a valuation that could reach $2 trillion. Beyond the headline number, here's what I'm most curious to learn.
When Anthropic filed confidentially for its IPO back in June, I wrote that it was one of the most important moments in the AI story so far. Now it's getting real. The company has reportedly picked the Nasdaq, marketing to investors could begin as soon as mid-October, and the listing is reportedly targeted for November. If it prices anywhere near the numbers being floated, it won't just be the biggest AI IPO ever. It could be the biggest IPO, period.
I use Claude, Anthropic's AI model, every day. It's the engine behind CIA, the portfolio management system I built and wrote about earlier this year, and I've watched it do in an afternoon what would have taken a team weeks not long ago. So I'm impressed by the product. But being impressed by a product and understanding a business are two very different things. This is a moment for curiosity, not hype.
How we got here
The growth story is almost hard to believe. Anthropic's Series A in 2021 valued the company at roughly $623 million. In May of this year, it raised $65 billion at a $965 billion valuation. That's more than a 1,500x increase in five years. Reports now suggest investors are targeting a public valuation around $2 trillion, with a raise that could match or exceed the record SpaceX set earlier this year.
The market backdrop is interesting too. IPO activity came in below expectations last quarter, with worries about AI spending among the reasons. OpenAI, meanwhile, has said it won't go public this year at all. Anthropic is stepping into a window its biggest rival just stepped away from. That's either confidence or opportunity, and probably both.
"For the first time, we'll get to see what it actually takes to build frontier AI."
What I'll be reading in the S-1
Right now, nobody outside the company and its bankers has seen audited numbers. That changes soon. Anthropic is required to make its financials public at least 15 days before its roadshow begins, and reports had pointed to that document landing in late September, so it could surface any day now. Here's what I'm most curious to learn.
The quality of the revenue. Bloomberg reported in August that annualized revenue had topped $65 billion, and investor projections reported since then put it above $100 billion by year end. Those are staggering numbers, but I care less about the headline than what's underneath it. How much is enterprise versus consumer? How concentrated is it among a handful of giant customers? How fast is it growing, and is that growth accelerating or starting to flatten? Those details say a lot about how deeply AI is actually being woven into how businesses operate.
The cost of running the machine. AI is expensive to build and expensive to serve. Every answer a model gives costs compute, and compute costs money. Anthropic has been lining up massive infrastructure deals and, according to Bloomberg, a $15 billion revolving credit facility, six times the size of the one it had last year. I want to see gross margins, how they've trended, and how much of the future is already committed to data center spend. Anthropic reportedly frames its total opportunity at more than $30 trillion. That's a huge vision, and the cost side of the ledger will show what it takes to pursue it.
The governance structure. Anthropic is a public benefit corporation, which means it's legally allowed to weigh goals beyond shareholder returns. Recent reporting on the draft filing adds another layer: the co-founders would hold majority voting power through a single special share, while an independent trust keeps the right to pick most of the board. Founder control isn't new in tech, but pairing it with a public benefit mandate is. I want to see how the final filing explains that balance to public shareholders.
The competition. Cheaper open-weight models are pushing some businesses toward options that cost a fraction of the top-tier price. Frontier quality has been Anthropic's edge. It'll be interesting to see how the company describes that landscape in its own words.
Why this matters beyond one company
This listing is going to be a milestone for the entire AI industry. For years, the inner workings of frontier AI labs have been almost completely private. A public filing forces a level of transparency we've never had before, and every company building with this technology will learn something from it.
My view hasn't changed since June. We're watching one of the defining companies of this era step into the spotlight, and that's worth paying close attention to. Whatever happens on listing day, the bigger story is what this moment says about how far AI has come, and how fast it's still moving.
"The companies that define an era eventually step into the light. When they do, we all get a clearer picture of where the future is headed."
— Paul Gravette, CEO & Founder, Gravette Capital