OpenAI Launches 'Useful Intelligence per Dollar' Framework to Redefine How Businesses Measure AI Value

OpenAI launches a 'Useful Intelligence per Dollar' framework, offering businesses a four-dimension scorecard to measure AI value beyond token costs, debuting alongside GPT-5.6, which sets a new coding benchmark record while using 54% fewer output tokens than competing models.

• OpenAI is introducing a new framework called 'Useful Intelligence per Dollar' to help business leaders measure the true value of AI spending beyond simple token costs.
• The scorecard evaluates AI performance across four key dimensions: the volume of useful work completed, the full cost per successful task, the dependability and accuracy of AI outputs, and whether value scales efficiently as usage grows.
• GPT-5.6, released with three performance tiers, demonstrates this approach in action, achieving a new state-of-the-art score on a leading coding benchmark while using 54% fewer output tokens than a competing model.