OpenAI’s Annualized Revenue Run Rate Exceeds $40 Billion
Bloomberg’s Rachel Metz reports that OpenAI is on track to exceed a $40 billion annualized revenue run rate, roughly twice its level at the end of 2025, driven by paying ChatGPT users, growing business adoption and its Codex coding assistant. She says consumer use remains OpenAI’s largest business, but the company’s rapid growth does not resolve its central economic problem: compute is both its biggest expense and an ongoing capacity constraint.
Demand is rising, but the economics remain unresolved
Tim Stenovec says sources expect OpenAI’s annualized revenue to exceed $40 billion—roughly double its run rate at the end of 2025. The figure is a run rate based on current performance, not a reported full-year revenue total. It indicates materially larger demand for OpenAI’s products, but does not establish what that demand produces after the cost of serving it.
That distinction matters because the services are compute-intensive. Rachel Metz says OpenAI has accumulated what she calls a war chest of computing power, yet still identifies obtaining enough compute as its biggest obstacle. Greater usage can lift revenue while increasing the amount of expensive computing capacity required to deliver the service.
Compute is really expensive. It costs a lot.
Metz says compute will consume a huge chunk of OpenAI’s revenue and is probably its largest expense by far. She did not provide specific spending, cost-of-revenue, or bottom-line figures. The $40 billion-plus run rate therefore establishes growth in commercial demand, not profitability.
A Bloomberg graphic cited a growing focus by OpenAI and Anthropic on a metric they say more accurately reflects the real cost and value of using AI software. It situated that effort amid pricing pressure from rivals in the United States and China. The emphasis reflects the unresolved issue behind the revenue figure: expanding sales and controlling the cost of delivering AI services are separate challenges.
Consumer scale leads, while business and coding broaden demand
According to Rachel Metz, OpenAI’s consumer operation remains “by far” its largest business. The company recently announced that it had passed 1 million active weekly users, a milestone it had been pursuing for a long time, and Metz says it has more paying users on the consumer side.
The revenue increase is not solely a consumer story. Metz also points to more businesses using OpenAI’s products, with Codex, its coding assistant, contributing to that adoption. Paying consumer users, business use, and coding-related demand together help explain why annualized performance is substantially above its late-2025 run rate.
The company’s consumer business is still by far its largest business.
The available account leaves the composition of that revenue unclear: Metz gives no figures for consumer subscriptions, business customers, or Codex, and no cost figures for serving each group. What is clear is the direction of the business. Consumer use remains the largest commercial base, while enterprise adoption and coding products extend OpenAI’s sources of demand beyond it.
Codex is growing in a market where Anthropic is strong
Rachel Metz resists reducing the market to a simple OpenAI-versus-Anthropic contest, stressing that many companies operate across the AI ecosystem. But she describes Anthropic as a meaningful rival that has proved itself over the last year or so.
Coding is where Metz draws the clearest competitive distinction. Anthropic has, “in some ways,” outpaced OpenAI in coding. That puts Codex in a more complicated position than a straightforward revenue-growth driver: it is part of OpenAI’s expanding business use, but it operates in a category where a close rival has particular strength.
OpenAI’s rising run rate nonetheless indicates that customers are willing to pay for its products, Metz says. Its growth does not depend on having a settled lead in every product category. Consumer scale remains the company’s largest business, while business adoption and Codex broaden the demand base; neither the revenue mix nor the cost structure behind that expansion is disclosed here.