OpenAI has reopened its $200 monthly Pro subscription to new customers, but the company is simultaneously cutting API credits in half as part of a broader shift away from flat-rate pricing toward usage-based billing models.
The move reveals OpenAI's strategy to transition users from subsidized subscription plans to pay-per-use arrangements. By reopening the Pro tier after a period of closure, OpenAI signals demand exists for premium access. But the halved API credits tell a different story: the company is reducing the value proposition to push subscribers toward consumption-based pricing.
OpenAI employee Thibault Sottiaux justified the reduction by pointing to efficiency gains in newer models. GPT-6 Sol and Luna, according to Sottiaux, deliver sufficient performance improvements that users get comparable results while consuming fewer API credits. The premise relies on customers accepting lower token allowances because better models require fewer tokens to accomplish the same tasks.
This strategy reflects OpenAI's evolving business model. Flat-rate subscriptions create ceiling effects. Once a customer pays $200 monthly, OpenAI loses direct incentive to optimize for their actual usage patterns. High-volume users subsidize low-volume users. Conversely, pay-per-use pricing aligns costs directly with consumption, allowing OpenAI to capture more revenue from power users while charging lighter users less.
The reopening of Pro signifies customer demand remains strong despite previous closure. OpenAI likely discontinued new signups during periods when server capacity couldn't meet demand or when the economics didn't work. Reopening now suggests the company has resolved infrastructure constraints or determined that managed demand growth justifies expansion.
The credit reduction operates as a gentle nudge. Users who complain about fewer credits will discover that updated models perform better per token. This reduces friction compared to a hard cutoff. Users comfortable with their current usage patterns face a choice: accept lower credits or migrate to metered billing where they pay only for what they consume.
OpenAI's shift reflects broader industry movement. Most AI platforms have moved toward usage-based pricing as their primary model. Flat rates work well for early adoption phases when usage patterns are unpredictable, but they create misaligned incentives at scale. Usage-based pricing provides predictability for customers, flexibility in spending, and better revenue capture for providers.
The timing matters. GPT-6 Sol and Luna's launch alongside the Pro reopening allows OpenAI to present a positive narrative: better models justify reduced token allocations. Without these efficiency improvements, cutting credits would appear purely extractive. Instead, OpenAI frames it as an upgrade path where customers get more capability per token.
This restructuring also affects OpenAI's competitive positioning. Anthropic, Google, and other AI providers all use usage-based pricing primarily. By moving Pro subscribers toward consumption-based models, OpenAI reduces differentiation between subscription and standard API access. The main remaining distinction becomes priority access and support rather than a fundamentally different pricing structure.
For Enterprise customers, this shift has little impact. They operate on custom contracts already structured around usage or compute commitments. For individual developers and small teams, the change forces a calculation: does the Pro subscription at $200 monthly still make sense, or does pay-per-use better match their actual consumption?
OpenAI's approach balances multiple goals: capture revenue from high-volume users, reduce subsidies implicit in flat-rate plans, encourage migration to usage-based pricing, and maintain a premium subscription tier for users who value priority features and support.