In short : Anthropic & OpenAI both re-rated their run rates in 2026 by segmenting : a mandatory enterprise repricing against an 80% price cut on the cheapest tier.

Like two SailGP boats in San Francisco Bay, Anthropic & OpenAI are vying to be the next multi-trillion public company & adding complexity to their strategies beyond technical one-upmanship.

Technology innovations marked the pre-2026 era : thinking models, bigger models, RL environments, agents, harnesses.

This year, business model innovation is more important, evident in two waves of the data. Both are segmentation moves : the same models, aimed at different buyers & priced to match. The first was Anthropic launching enterprise metered billing in March of 2026, which doubled revenue in a quarter.

About three months later, OpenAI responded with a business model innovation of its own, cutting the price of Luna (its most affordable model) by 80%. The move has catapulted OpenAI to within a boat length.

OpenAI is approaching $70b in run rate. Margins remain ambiguous & gross profit dollars will likely be a better way of evaluating the relative strength of the businesses.

Some of these strategic changes impact revenue at this level of scale in a single quarter, an indication that the market is still fluid. Each of these businesses will near $100b in revenue by end of year.

Yesterday’s revelations from the leaked Anthropic S-1 suggest even the largest buyers of AI haven’t yet chosen. Two customers, Amazon & Google, comprised nearly a quarter of Anthropic’s revenue last year, & neither is locked into a long-term contract.

Technical innovation, customer segmentation, price discrimination : we are watching a business school case play out on the water.