Jensen Huang just dropped a line that, stripped of context, could send crypto Twitter into a frenzy. During Nvidia’s fiscal Q1 2027 earnings call, the CEO declared that “tokens are now profitable” for AI companies. Before anyone starts minting celebratory NFTs, here’s the thing: he’s talking about AI output tokens, not blockchain tokens.
The distinction matters enormously. In AI, a “token” is a unit of model output, a chunk of text, code, or reasoning that a large language model generates. Huang’s point is that producing these outputs has crossed a profitability threshold, meaning AI companies can now charge more for their model outputs than it costs to generate them.
Nvidia’s numbers tell the real story
Nvidia reported quarterly revenue of $81.6 billion for Q1 FY2027, an 85% increase compared to the same period a year earlier. The data center segment, which houses the GPU infrastructure powering most of the world’s AI workloads, brought in $75.2 billion. That’s a 92% year-over-year jump.
Huang described the current demand environment as having “gone parabolic.” His exact framing: “Tokens are now profitable. So model makers are in a race to produce more.” In other words, the economics have flipped. AI inference, the process of running trained models to produce useful outputs, is no longer just a cost of doing business. It’s a revenue engine.
