OpenAI launched its Decisions API on October 22 to help AI agents make quick, low-cost choices. This move puts the company in competition with TypeSafe’s Jev. OpenAI product lead Nikunj Handa said, "decision times had come down from between one and three seconds to between 100 and 180 milliseconds."

Synopsis

The Decisions API, now generally available, is meant for tasks where an AI agent only needs to choose between a few options. It puts OpenAI in competition with decision models such as TypeSafe's Jev.

OpenAI has launched an application programming interface (API) that lets AI agents make quick, low-cost choices from a set of predefined options. The launch puts the company in competition with TypeSafe's Jev, one of the first AI models built for such tasks.

The Decisions API, which became generally available on October 22, is meant for tasks where an AI application does not need a detailed answer from a model and only needs to pick between a few options.

Every time an AI agent classifies an input, chooses a tool or decides its next step, it makes a separate call to a model. When a large model handles each of these small decisions, the time and cost add up across an entire workflow.

What are decision models?

Jev helped popularise this category, known as decision models or System One models. The term is borrowed from psychology, where it refers to fast, instinctive thinking. Unlike general-purpose models, which generate detailed responses, decision models focus on quick choices and can give a confidence estimate for each decision.

Live Events

How is OpenAI's approach different?

Using a general-purpose model such as OpenAI's GPT-6 Luna to decide whether a customer query should be escalated, or which team should handle a request, is more than the task needs.

With the Decisions API, developers define a question and the possible answers. The API returns a structured choice, score or probability, which the application uses to decide what happens next.

Instead of building a separate decision model, OpenAI has reconfigured GPT-6 Luna to make these choices faster and at a lower cost. The company says the API can make decisions up to 10 times faster than using Luna through its Responses API.

Nikunj Handa, OpenAI's product lead for APIs, told TechCrunch that decision times had come down from between one and three seconds to between 100 and 180 milliseconds.

OpenAI charges $0.10 per million input tokens, the units of text a model processes, and $0.01 per million for cached inputs. There is no charge for output tokens.

Where is it being used?

The use cases include classifying financial transactions, routing customer support requests and making language-learning apps more interactive. OpenAI already uses the API in its Dots voice assistant to respond faster, and its personal finance team uses it to categorise transactions.

The API also works with OpenAI's GPT-Live voice system. A customer support application, for instance, could use it during a live conversation to decide whether to ask a follow-up question, send a request to another team or escalate an issue.

Speak, an AI-powered language-learning platform and an early customer, told TechCrunch that it had cut the time between actions in its app from 1.4 seconds to 0.13 seconds.

Can OpenAI compete with Jev?

Speed and cost are only part of the equation, as a decision model also needs to make the right choice consistently. Benchmarks cited by TechCrunch show that models from Liquid and Base Lab have performed strongly on decision-model evaluations, while OpenAI's offering ranked lower on two indices and Jev remained competitive.

For developers, the real test will be how well these models handle their own tasks, including unclear inputs.

KYB and KYC: Business behind the wealth creator called ‘Divi’s Lab’ – Part 2, concluding

KYB and KYC: Business behind the wealth creator called “Divi’s Lab” – Part 1

The global business of promising more

One company, 2 words. Bond yield and CDS: Will they decide the fate of global markets?

Business of Odd Things: Inside an industry that farms fish that nobody is meant to eat

War's second-order effects are becoming visible in critical sectors