OpenAI's Enterprise Revenue Just Crossed $12B. The Real Story Is Margin.
Inference costs are collapsing faster than pricing. That changes the entire moat conversation.
Inference costs are collapsing faster than pricing. That changes the entire moat conversation.

Illustration by IMF Alpha editorial · Reviewed by Priya Raman
OpenAI's enterprise book of business hit a $12 billion annualized run-rate this quarter, according to two people familiar with the matter — a milestone that, on its own, would make it one of the fastest-scaling software companies in history.
But the more interesting line in the model is gross margin. Cost-per-million tokens for the company's flagship reasoning model has fallen roughly 71% year-over-year, while enterprise contract pricing has compressed only 18%. The delta is now wide enough that OpenAI's marginal contract economics rival mature SaaS comparables.

Recent Federal Reserve hawkish signaling has initiated a re-evaluation of growth technology stock valuations, creating a potential disconnect between market sentiment and long-term prospects.

Regulatory bodies are increasing scrutiny of artificial intelligence in financial markets, focusing on risk management and transparency in automated trading systems.

As enterprises shift from chasing bigger models to buying better data, new marketplaces are rewriting the rules for chips, cloud costs and startup valuations.