YC CEO Backs Tokenmaxxing to Help AI Startups Scale

Y Combinator Chief Executive Garry Tan is championing “Tokenmaxxing,” an approach that treats spending on AI tokens and agents as an investment in future productivity rather than merely an operating cost. The idea matters for early-stage companies because capable AI agents could allow a handful of employees to perform work that previously required much larger teams, helping startups accelerate product development and compete with better-funded incumbents.

Tan urged founders to “Skillify” successful agent workflows by converting them into explicit, reusable instructions that can be executed repeatedly and improved over time. He argued that this accumulated operating knowledge could help very small startups build products faster and generate substantial recurring revenue. The cited report did not specify when Tan issued the recommendation, how much founders should spend on tokens, or a dollar figure for recurring revenue achieved through the strategy.