# Jev

Jev is TypeSafe AI's decision model, launched September 15, 2026, and the first sold as a
System One model. It doesn't generate text. You send it state, text or JSON, and named
questions: a choice between options you define, a score on ordered levels, or a yes-or-no
probability. Each comes back as a typed answer with a probability and a confidence score.

Questions run independently and in parallel, so you split a judgment into narrow questions
and combine the answers in code. Jev isn't fine-tuned per customer; everything you control is
in the request. It's trained for calibrated probabilities rather than preferred text. An early
independent test found its choices and scores overconfident, most of all on questions the
input can't answer.

TypeSafe's own numbers put it at 70 to 500 milliseconds per request and $0.042 per million
input tokens. Each version ships with a published list of how it fails, including literal
reading, counting, and injected instructions.

## Questions answered under this concept

- [Can I self-host Jev?](https://tessary.ai/answers/jev/can-i-self-host-jev)
- [Can Jev choose which tool an agent calls?](https://tessary.ai/answers/jev/can-jev-choose-which-tool-an-agent-calls)
- [Can Jev replace an LLM judge for grading agent traces?](https://tessary.ai/answers/jev/can-jev-replace-an-llm-judge)
- [Does Jev beat a frontier LLM's accuracy on the same task?](https://tessary.ai/answers/jev/does-jevs-accuracy-beat-a-frontier-llms)
- [How do I test Jev against the LLM call it would replace?](https://tessary.ai/answers/jev/how-do-i-test-jev-against-the-llm-call-it-would-replace)
- [Is Jev more consistent than an LLM judge on the same trace?](https://tessary.ai/answers/jev/is-jev-more-consistent-than-an-llm-judge-on-the-same-trace)
- [Is Jev safe to run on untrusted user input?](https://tessary.ai/answers/jev/is-jev-safe-on-untrusted-user-input)
- [What is Jev?](https://tessary.ai/answers/jev/what-is-jev)
- [Which agent tasks is Jev good for, and which should stay on an LLM?](https://tessary.ai/answers/jev/which-agent-tasks-is-jev-good-for)

---

Source: https://tessary.ai/answers/jev
All concepts: https://tessary.ai/answers
From Tessary, agent reliability for AI agents in production: https://tessary.ai