In late 2024, OpenAI priced at 46× annual revenue ($157B / $3.4B). Anthropic at 41× ($61.5B / $1.5B). xAI's most recent primary works out to over 300× revenue, though revenue is barely defined yet.
These are extraordinary multiples by any historical standard. The question employees should ask is: are these multiples a structural reset (because AI is genuinely different) or a temporary peak (because of capital flowing in)?
Public peer comparison
- Nvidia: 28× revenue, 85% growth. The market is paying for both growth and dominance.
- Microsoft: 12.5× revenue, 17% growth.
- Palantir: 42× revenue, 30% growth. The closest public peer to OpenAI by multiple.
Public AI-adjacent companies generally trade at 12–28× unless growth is extreme (Palantir) or there's a monopoly story (Nvidia). Private AI labs trading at 40–50× either need extreme growth or extreme dominance.
The historical comparable
Only the most extreme dot-com IPOs — Yahoo, Amazon at the very peak — traded at 30–50× revenue. Most of those compressed by 80% over the following two years. Survivor bias means we remember the ones that grew into the multiple.
What it means for employees
If you're at a top-tier AI lab today, your equity is priced for near-perfection. Slight growth deceleration or a competitor breakthrough could compress your paper wealth by half overnight. Don't make life decisions assuming today's valuation is the floor.
Practical implication: if you have a meaningful position, partial liquidity (via tender or secondary sale) at today's prices is mathematically attractive — even at a 10–20% discount, you're locking in extraordinary realised gains.