The fastest growing companies in AI & software all share one property. They are either selling AI directly, or they are reselling inference. At worst, the first derivative of inference.
State-of-the-art AI email costs $22-130/month, but smaller models & local inference can cut costs by 100x. The next 12-24 months will be defined by matching models to workloads.
Join our 2026 GTM survey analyzing how startups adapt sales, marketing & AI. Get exclusive data on efficiency trends, AI deployment models, and the emerging impact of AI buyers.
About half of agent tasks can run on a local 35B model. The real advantage isn't cost or privacy -- it's latency. 2.1x faster means more iteration cycles per session.
Anthropic trades at 17x EV/NTM on 165% growth, a 65% discount to Palantir. Capital intensity, profitability uncertainty & growth volatility explain the gap.