OpenAI CEO Sam Altman revealed that the company's heaviest token consumer now uses approximately 100 billion tokens per month, a staggering 1 million-fold increase from six and a half years ago when the top user consumed just 100,000 tokens monthly. Speaking at an enterprise event on Tuesday, Altman highlighted how dramatically AI consumption patterns have shifted, with what was once considered extraordinary usage now approaching average levels across the industry.

How Has Token Consumption Evolved at OpenAI?

The scale of growth in AI token usage tells a remarkable story about how quickly artificial intelligence has moved from experimental technology to essential business infrastructure. Altman noted that six and a half years ago, OpenAI's top token consumer was "very likely the token leader in the world" at 100,000 tokens monthly. Today, that same consumption level represents roughly the per capita average globally, demonstrating how normalized heavy AI usage has become.

Altman

What makes Altman's revelation even more striking is that OpenAI's internal top user is not actually the world's heaviest token consumer. Altman acknowledged finding external users who spend more tokens than OpenAI's own record holder, calling it a personal "embarrassment." This admission underscores how rapidly AI adoption has accelerated beyond the companies building the technology themselves.

Why Are Token Costs Becoming a Major Business Concern?

The explosive growth in token consumption has created an unexpected crisis for enterprises. According to reporting on OpenAI's internal culture, the company maintains a token leaderboard where employees sometimes publicly post their consumption totals on social media. Some individual contributors have reached extraordinary levels; one OpenAI employee spent 210 billion tokens in a single week, while creator Peter Steinberger of OpenClaw spent $1.3 million worth of tokens in one month, consuming 603 billion tokens in 30 days.

This internal enthusiasm for token spending contrasts sharply with how other major technology companies are now responding to rising AI costs. The shift from celebration to cost-consciousness happened remarkably quickly. Altman noted that the question of cost "came rather quickly," and that the issue never came up at the beginning of 2026 when "people were totally happy with the amount they were spending." Now, AI costs represent "a huge issue" for organizations across the industry.

Altman

Steps Companies Are Taking to Control AI Spending

• Uber's Spending Caps: The ride-sharing company has instituted a new rule placing a monthly $1,500 cap per employee and per agentic coding tool, including tools like Anthropic's Claude Code and Cursor, to prevent unchecked AI expenses.
• Amazon's Leaderboard Removal: Amazon has reportedly scrapped its internal AI leaderboard as costs soared, with senior executives instructing staff not to "use AI just for the sake of using AI" to reduce wasteful consumption.
• OpenAI's Efficiency Focus: Despite maintaining its own token leaderboard culture, OpenAI is continuing to push its models forward while exploring alternative ways to deliver "more value for less spend" to address customer concerns.

The contrast between OpenAI's internal token celebration culture and the broader industry's cost-cutting measures reveals a tension at the heart of AI adoption. While OpenAI employees and external power users have embraced massive token consumption as a sign of innovation and capability, enterprise customers are increasingly treating AI spending as a budget line item requiring strict governance.

Altman referenced a popular industry meme that captures this moment perfectly: "My company spent my entire 2026 budget in Q1, can you make this more efficient?" This joke reflects a genuine pain point that has emerged across enterprises experimenting with AI tools. The rapid acceleration from unlimited enthusiasm to urgent cost management suggests that organizations are still learning how to integrate AI into their operations sustainably.

The token consumption explosion also raises questions about the sustainability of current AI pricing models and whether the industry's cost structure can support the scale of adoption that vendors are promoting. As more organizations hit budget constraints and implement spending caps, AI providers like OpenAI will face pressure to improve efficiency and offer more cost-effective solutions to maintain growth momentum.