OpenAI Charges Some Large Clients Based on Task Completion

According to The Information, OpenAI has begun offering some large clients a new payment model in recent months, where customers pay only after the AI actually completes a task. Implemented use cases so far include customer service, where enterprises can pay based on AI-completed service tasks rather than purely on tokens, call volume, or seat count. OpenAI has not disclosed specific clients, unit pricing, or the criteria for "task completion," and declined to comment. This is not yet a new package available to all enterprises; the public pricing for Enterprise on the official website still primarily charges based on quotas and tokens, making this outcome-based pricing more of a separately negotiated contract for select large clients. In January, OpenAI CFO Sarah Friar suggested that future AI business models would adopt outcome-based pricing. In July, OpenAI further proposed that, rather than token unit prices, the focus should be on the total cost of getting the AI to complete a task—an approach that is now beginning to enter real contracts.