AI Certificates

System One models

Jev by TypeSafe AI: the decision model, explained

Matthew Hartman · CCAR-P certified, scored 965/1000 · 7 min read

Every model you have used was built to write. Jev gave that up on purpose: it decides, in under half a second, and hands your code a number saying how much to trust the decision. This is what that buys you, what it costs, and how to tell which half of a problem belongs to it.

Practice this on Udemy

Jev by TypeSafe AI: Practice Test for Engineers

Two timed tests, 60 scenario questions, every option explained — including the three you did not pick. $19.99 on Udemy.

Jev shipped into early access on 15 September 2026, which means nothing about it is in any model's training data and most of what is written about it is guesswork. Everything below is written against TypeSafe's published documentation.

The trade

Every model you have used was built to write. That single capability is what makes language models feel general, and it is also what makes them slow, expensive and hard to fully trust.

Jev gave it up. It cannot write you a sentence. What it does instead is decide — in 70 to 500 milliseconds, returning a value from a set you defined, with a calibrated confidence number your code can gate on.

The name for this is a System One model. Psychologists split human thinking into two systems: System Two is deliberate and generative, the part that sits down and reasons something out; System One is the fast, automatic part that recognizes a face or judges that a room is tense. Language models were built for System Two, and we have been asking them to do System One work because they were the only tool on the table. That is why an agent spends twelve seconds and real money answering "is this ticket urgent."

The question to ask is never "can Jev do this?" It is "is this half of the work a decision?" Almost every agent call has a decision half and a generation half tangled together. Finding the seam is the whole skill.

The three primitives

Pick by what the answer means, not by what feels natural to write.

TakesReturnsUse when
ChoiceOptions plus stateThe chosen option, a probability for each, and a confidenceThe alternatives are distinct and unordered
ScoreState plus 2–10 ordered levelsA probability-weighted value, probabilities per level, a legend, a confidenceYou are measuring a position on a spectrum
NoulA statement plus stateA probability from 0 to 1 that the statement holdsThe question is genuinely yes or no and you will threshold it

Picking wrong does not raise an error. It returns a plausible, well-formed, confidently-scored answer to a question you did not mean to ask, which is the failure worth designing against.

The test for Choice versus Score: shuffle your options. If shuffling changes nothing — billing, technical, account, spam is the same set in any order — it is a Choice. If shuffling destroys the meaning — low, medium, high, critical has a direction — it is a Score, and using Choice throws away the ordering.

The test for Noul versus a two-option Choice: ask what your code does next. If you are going to compare against a threshold, you want a Noul. If both labels are real categories, refund versus exchange, use Choice.

A Score is a position, not a level

The most misread thing in the API. A Score does not return one of your levels. It returns each level number multiplied by its probability, summed — so a four-level rubric can come back as 2.6, and there is no level 2.6. Round it to the nearest level and you have thrown away exactly the information you called Score to get.

Confidence is the reason the model exists

Every Choice and Score answer carries a confidence between 0 and 1, derived from how the probability mass is spread across your options or levels. It adds no information the probabilities did not already carry; it is a convenient summary of the distribution's shape, which is why TypeSafe also returns the full probabilities so you can compute your own measure if theirs does not fit.

A Noul carries no confidence. It returns one probability, and how far that sits from 0.5 is the signal. An audit or a gate keyed on confidence silently skips every Noul in a request — usually where the guardrails are.

The documented way to use it is three tiers, not one number:

Where you draw the lines depends on what happens when you are wrong. Read-only work can act at moderate confidence. Destructive operations want above 0.9. The advice is to start conservative, test against your own data, and adjust.

A gate that never fires is not broken. If your task is genuinely easy, the probability mass really is peaked and high confidence is the honest answer. Raising the bar until something trips manufactures uncertainty that is not there, and you will start routing correct answers to humans to justify the check.

Where it earns its keep

Three patterns show up first in production, and all three are the same move: a bounded decision placed in front of unbounded generation.

The design principle underneath all of them: decompose. Ask atomic questions in one request rather than one compound question. Every question in a request is evaluated in parallel against the same state, so four questions cost roughly what one costs, and the combination logic lives in your own code where you can read and test it.

Where it is the wrong tool

The numbers, and the ones that bite

Early access means these can move. Check the docs before you build a design around any specific figure.

Check whether it stuck

Reading about a decision model and being able to use one are different things. The questions below are sixty scenarios — situations an engineer would actually be in — with every option explained, including the ones you did not pick.

Jev practice tests on Udemy — 60 scenario questions, every answer explained

Two timed tests. The first covers the decision-or-generation split, the three primitives and confidence. The second covers the documented patterns, state shapes and the limits above.

There is no Jev certification, and anyone selling you one is inventing it. This is practice, which is a different and more honest thing.

Common questions

What is Jev by TypeSafe AI?
Jev is TypeSafe AI's first System One model, in early access since 15 September 2026. It turns natural language and application state into a typed decision — one option from a set you defined, a score on a rubric you wrote, or a probability that a statement holds — in 70 to 500 milliseconds, with a calibrated confidence attached. It does not generate text.
What are Choice, Score and Noul?
The three question types. Choice picks one option from a defined set and returns the choice, a probability for every option, and a confidence. Score rates the state against two to ten ordered levels and returns a probability-weighted value that can land between levels. Noul takes a statement and returns the probability it holds — and carries no separate confidence.
Is Jev a replacement for an LLM?
No. It is the other half. Jev handles the bounded decision; a language model handles anything that ends in prose. Most agent calls have both halves tangled together, and the skill is seeing the seam.
What does Jev cost?
$0.042 per million input tokens. Output tokens are free, because there are none — the model returns a typed value rather than generated text.
What are Jev's limits?
Text input only, so images must be pre-processed. Single-stage selection caps at 255 options, and beyond that you go two-stage. Score levels run from two to ten. Context is 64k per request, with 32k covering the state plus the longest single question. English is the primary training language.
Is there a Jev certification?
No. TypeSafe has not announced a certification, and nobody can sell you one. What exists is practice: this guide, and a set of scenario questions on Udemy for checking whether the ideas here actually stuck.

Practice this

A full-length practice form for every one of these is free here, with every answer option explained, including the wrong ones.

Every set beyond the free forms, across all four certifications, in one payment.

Get all four $79