OpenAI Makes Its Decisions API Model Widely Available to All Customers to Cut Costs

OpenAI has made its Decisions API model widely available to all customers, entering a growing market for fast and inexpensive intelligence tools. The move follows the rise of startup competitors like TypeSafe, which popularized deep learning models that act as super classifiers to make rapid choices with confidence ratings.

Announced on September 29 at OpenAI’s developer day and rolled out broadly on Thursday, the tool was built by reconfiguring the ChatGPT Luna model to cut costs and speed up inference. Product lead Nikunj Handa stated that response times dropped from one to three seconds down to between 100 and 180 milliseconds.

The public release also includes new voice and visual modalities, alongside geographic targeting to reduce global latency. Early enterprise testers, including Box, Datadog, Ramp, and Speak, used the API to replace standard large language models with faster alternatives. Speak CTO Andrew Hsu reported that response times dropped from 1.4 seconds to 0.13 seconds.

Meanwhile, industry benchmarks remain varied and competitive, with models from Liquid, Base Lab, and TypeSafe frequently trading top spots across different evaluation indexes.