Artificial Intelligence

OpenAI makes its Jev competitor widely available

OpenAI has made its Decisions API model widely available to customers, entering the market for fast, cheap intelligence that recently elevated startup TypeScale into a multi-billion dollar valuation.

Radwan AbdulhadiEditor and technology writer

Published: October 23, 2026 · Updated: October 23, 2026 · 3 min read

OpenAI makes its Jev competitor widely available — Image: TechCrunch

OpenAI Expands Decisions API

OpenAI has made its Decisions API model widely available to users for the first time, officially entering the market for fast and cheap intelligence. This market segment recently made TypeSafe—a startup founded by a former researcher at the frontier lab—a $7.6 billion company.

TypeSafe’s Jev model originally created a category for deep learning models that act as super classifiers. Given a set of options, these models can quickly make choices and offer a confidence rating in their decision. Software developers quickly discovered that these System One models could replace LLMs in various workflows, operating faster and at a dramatically lower cost.

Development and Technical Approach

Nikunj Handa, the product lead for OpenAI’s APIs, noted that his colleagues saw Jev and immediately recognized numerous use cases. Rather than building a specifically calibrated decision model from the ground up, Handa’s team took ChatGPT Luna and reconfigured it to speed up inference and lower costs.

“The big unlock with Decisions API has been the speed, shrinking things from one to three seconds to 100, 120, 180 milliseconds,” Handa stated, explaining that plain old Luna gets the job done for now while future training may introduce further enhancements.

Customer Implementation and Latency Reductions

Initially announced at the company’s developer day on September 29, the Decisions API became available to all customers. OpenAI integrated new voice and visual modalities into the public version and targeted its inference resources geographically to ensure lower latency worldwide.

In initial trials, engineering teams at companies such as Box, Datadog, Ramp, and Speak utilized the API to replace OpenAI’s traditional LLMs with higher throughput. Andrew Hsu, CTO at AI language tutor builder Speak, reported that the Decisions API helped cut action turnaround time from 1.4 seconds down to 0.13 seconds.

Benchmarks and Competitive Landscape

Evaluating decision models remains complex, as optimization across intelligence, cost, calibration, and speed varies by workload. TypeSafe CEO Diogo Almeida encourages users to trial models in their own work, though the company also released its own rankings alongside the rollout of its latest Jev model.

Public evaluation frameworks like the AIM System One Index and the Artificial Analysis Decision Index currently lean toward Liquid's D1 Ultra and Base Lab's Elephant-1. Meanwhile, the latest Jev model maintains competitive positioning, while TypeSafe's offerings remain the most widely-queried options on OpenRouter.

Unique Testing and Future Outlook

One notable evaluation method is JevMan, a benchmark where decision models play PacMan. Felix Wunderlich of Opper AI described the classic game as a rigorous test due to its nonstop stream of quick, high-stakes decisions requiring choices within hundreds of milliseconds.

As these tools become faster and more efficient, they align with economic principles highlighted by Jevons paradox, potentially unlocking a broader set of hidden use cases.