OpenAI Keeps Token Pricing After Developer Backlash
The company abandoned a move to opaque flat-fee billing after developers demanded transparency in how usage costs are calculated.
Updated on Oct. 29, 2026 in Artificial Intelligence
Live Poll
Do you prefer transparent itemized billing over opaque flat-fee models for your digital service subscriptions?
OpenAI recently evaluated shifting its billing model from token-based pricing to a flat, per-job fee structure similar to AWS. The organization ultimately decided to maintain current token-based billing following negative feedback from developers who prioritized cost transparency.
Why it matters
Developers require granular insight into billing calculations to manage operational expenditures, making them resistant to opaque, flat-fee pricing models. This decision reflects the challenge of pricing non-deterministic AI outputs while maintaining clear cost signals for enterprise users.
Sam Altman has publicly characterized tokens as an inadequate metric for quantifying AI work. The company has navigated a tenfold annual decrease in usage costs, which complicates the shift toward stable, predictable pricing models.
The players
OpenAI
An AI research and deployment organization that develops large language models and provides them via API for third-party developers.
Sam Altman
The CEO of OpenAI who oversees the organization's strategic approach to developer tools, pricing, and long-term research goals.
The details
Under the proposed per-job model, developers would have submitted prompts and received a final, fixed price for the computation instead of being billed based on the volume of tokens—discrete units of text processed by the model. Developers rejected this approach because it obscured the underlying cost drivers, raising concerns about potential long-term financial disadvantages. The current token-based system allows users to track the complexity of inputs and outputs against their budget.
Timeline
October 29, 2026: Sam Altman discussed pricing strategies on the Cheeky Pint.
The Tech Race
The debate highlights a tension between simple, predictable pricing and the complex nature of AI inference. OpenAI remains aligned with token-based models, diverging from the fixed-unit pricing found in legacy cloud infrastructure.
Developers using the OpenAI API will continue to see usage-based billing measured in tokens rather than fixed per-job fees. This ensures that cost calculations remain transparent as usage scales or model performance changes.
The takeaway
The preference for transparent token pricing underscores how sensitive AI-integrated workflows are to cost predictability. Observers should track further comments from leadership regarding alternative units of measurement that may eventually supersede tokens as the primary billing mechanism.
Further reading
Explore more on the current landscape of Artificial Intelligence to understand how infrastructure providers are evolving.
Live Poll
Do you prefer transparent itemized billing over opaque flat-fee models for your digital service subscriptions?
