OpenAI’s $40B Revenue and the Token Billing Dilemma
OpenAI has reached an annualized revenue of $40 billion, yet it continues to sell model access by token rather than by outcome. This discrepancy has sparked a heated debate on Hacker News: if models are approaching human-expert capability, why doesn’t OpenAI simply launch 100 companies and capture the profits directly? The discussion intensified after Codex shifted to token-based billing, with enterprise API spending surging 82% quarter-over-quarter.
The Token Economy: A Double-Edged Sword
Token-based pricing has been a scalable way to monetize AI, but it creates a misalignment of incentives. Customers pay for usage, not results, which means OpenAI profits even when the model fails to deliver value. This model has fueled rapid revenue growth but also exposes the boundary of AI commercialization: the gap between model capability and reliable, autonomous business execution remains wide.
The 82% surge in enterprise API spending indicates that businesses are embedding AI deeper into workflows, but it also raises questions about cost predictability and ROI. As models become more capable, the pressure to shift from token-based to outcome-based pricing will grow.
Why Not Own the Companies?
Owning 100 companies would require operational expertise, capital allocation, and risk management far beyond building models. OpenAI’s core competency is research and product innovation, not running diverse businesses. Moreover, such vertical integration could stifle the ecosystem of developers and startups that build on its API, potentially reducing demand for its core offering.
However, the debate highlights a real tension: as AI agents become more autonomous, the value may shift from the model provider to the application layer. OpenAI must decide whether to capture that value through equity, revenue-sharing, or by enabling others to succeed.
Forward-Looking Perspective
The future of AI monetization likely lies in hybrid models: a base fee for access plus performance-based pricing for outcomes. OpenAI’s next move will signal whether it sees itself as a utility provider or a value creator. For now, the token billing model remains a powerful engine, but its limits are becoming visible. The company’s ability to evolve its pricing strategy will determine whether it can sustain growth as competition intensifies and enterprises demand measurable results.
