The Anthropic IPO ยท scenarios
What is Anthropic worth, on your view of AI?
Two things decide most of it: how fast AI progresses, and who ends up with the money. Move the two sliders and watch the price.
On your view, a share is worth
Every scenario, value per share
How this works and what is held fixed
Values. Each scenario's value per share comes from the valuation engine v7 (mkt/anthropic-valuation), through The Long Swell's values.js, unchanged. A slider between two stops blends the two scenarios in proportion.
Held fixed. Anthropic takes 45% of the AI labs' revenue. The labs' own revenue plans and the more cautious path (revenue tied to the jobs AI replaces) count half each; the AGI scenarios use the labs' plans only. A funding crash in 2027-29 is counted at the model's odds: 60% if AI stalls, 64% if it keeps improving, none if AGI arrives.
The four rows follow The Long Swell's four AI pathways. "Specialist AI firms win" merges two of the model's worlds, which use the same computing and differ only in who keeps the money: labs selling by the use (74%) and specialist firms owning the client (26%), at the model's own weights. On the model's own weights over all its scenarios, the average value is $36 a share. The page opens there: the speed slider at the model's weighted position (25% stalls, 70% keeps improving, 5% AGI) and the money slider where that gives $36.
Unverified. None of the scenario values is verified against outside data, and they were built on an earlier version of the compute model than the current one.
Research, not a recommendation to buy, sell or subscribe. Values rest on scenario-model figures that are not yet verified. The best AGI cases are upside cases, not forecasts. The model was built in a Claude session, and Anthropic makes Claude.