Anthropic Is Raising $50 Billion at a $900 Billion Valuation. Let's Talk About What That Number Actually Means.
The biggest private fundraise in history is happening right now. Here's the mental model you need to read it correctly.
Anthropic is considering a funding round of up to $50 billion at a valuation of roughly $900 billion. The company's annualized revenue run rate is expected to surpass $45 billion soon, up from $9 billion at the end of 2025. If that revenue trajectory holds, this is one of the most extraordinary growth curves in technology history.
A board decision is expected in May, and a potential IPO could come as early as October 2026. If the round closes at these terms, Anthropic would surpass OpenAI as the most valuable private AI company in the world.
Let's put the numbers in context. Anthropic raised at $380 billion in February, just three months ago. A $900 billion round would more than double that. A company adding $500 billion in implied valuation in 90 days isn't an incremental step. It's a category redefinition.
The mental model: Expected Value
Expected Value is one of the models in my book. The formula is simple: probability times payoff. The mistake most people make is treating high-conviction outcomes as certainties, and low-probability outcomes as impossibilities.
Applied here: investors bidding at a $900 billion valuation are implicitly pricing in a specific expected value calculation. They're saying the probability of Anthropic becoming a multi-trillion dollar business, multiplied by the payoff of getting in at this price, is positive. That calculation depends almost entirely on the revenue trajectory holding. Anthropic's revenue is believed to be closer to $40 billion run rate currently, driven heavily by Claude Code and enterprise adoption.
What makes this intellectually serious, not just hype: the revenue is real and growing fast. What makes it still a bet: a $900 billion valuation requires a company that doesn't just maintain that growth but accelerates it for years in a market where OpenAI, Google, Meta, and xAI are all competing with similar infrastructure.
The contrarian take isn't that Anthropic is overvalued. It might not be. The contrarian take is that most coverage treats the valuation as the story. The actual story is the revenue trajectory. A company going from $9 billion to $45 billion annualized in one calendar year has demonstrated something genuinely unusual. If the model holds for another 12 months, the current valuation looks defensible in retrospect. If the enterprise market saturates or a competitor closes the capability gap, it doesn't.
Expected value thinking says: don't anchor on the number. Anchor on the probability distribution of the outcomes that justify it.
I've been thinking about this framework constantly while building /mkt. The regulated stack we've built isn't a sure thing. It's a high-conviction bet with a specific expected value calculation behind it. That's how every serious company gets built.