Jev becomes the fastest-adopted model in AI Gateway history

TypeSafe AI's Jev crossed more than twice the paid-team count of any prior model launch within its first 24 hours on AI Gateway, according to the September AI Gateway Production Index. No previous launch was adopted as quickly.

Jev overtook every comparison model within twelve hours and kept expanding its lead through the rest of the day. At the 24-hour mark, nearly 13% of paid teams were using it — 2x the GPT-5.6 family and more than 6x Fable 5.1.

Jev reached a tenth of teams within 18 hours. Every other recent model launch remained below 7% after a full day.

The first day of adoption was the best of any recent launch, but whether that early usage holds is still an open question.

What Jev is

Announced on September 15, Jev is a probabilistic decision model aimed at structured decision-making within software. An application submits context plus a list of questions; Jev works through the questions in parallel and returns typed choices, scores, or true-or-false answers, along with probabilities.

Because those answers arrive in a format code can consume directly — rather than the text a general-purpose language model produces — developers can put Jev to work on tasks such as:

  • picking an agent's next tool or subagent
  • deciding whether a workflow should continue, retry, ask the user, or stop
  • scoring urgency or risk before an action is taken
  • verifying model outputs, enforcing guardrails, or routing uncertain cases to human review

TypeSafe AI's own workflow evaluations put Jev at up to 194 times faster and 445 times cheaper than language models.

The September AI Gateway Production Index has further model usage data, and Jev can be tried through AI Gateway.