Bessemer Venture Partners separates the AI-native company cohort into two groups with very different unit economics underneath top-line growth that looks comparable.

AI Supernovas: explosive growth, thin-wrapper products, gross margins around 25 percent.

AI Shooting Stars: disciplined operators, infrastructure-mature, gross margins closer to 60 percent.

The two groups are often valued similarly because their growth rates look comparable in the latest quarter. The unit economics are not comparable. The implications compound differently as growth slows.

The Shooting Stars will compound. The Supernovas will compress when the inference bill cannot be outrun with new ARR.

Top-line ARR has become a less useful indicator of AI-native company health than it was for SaaS. The unit economics matter more.