Guides
14 guides on getting real work out of Jev: what it is, where it wins, where it breaks, and how to build around the edges.
Fundamentals
What Is Jev? A System One Model, Explained Without the Marketing
9 min readJev returns typed decisions and probability distributions instead of text. Here is the mental model, what actually comes back, and who should care.
Jev Is Not an LLM, and That's the Whole Point
8 min readNo string output means parallel evaluation means free output tokens. The architectural tradeoff behind Jev, and where the story gets muddy.
'It Can't Hallucinate' — What That Claim Actually Means
8 min readType safety guarantees schema conformance, not factual correctness. What Jev genuinely rules out, what it doesn't, and why the distinction matters.
Understanding Confidence and Calibration
11 min readConfidence is derived from the probability distribution, not reported separately. How it's computed, why Noul has none, and how to set thresholds.
Engineering
Your First Jev Integration, End to End
12 min readcurl to Python to TypeScript: the request shape, reading answers correctly, error handling with backoff, and why you should pin the version.
How to Use Jev Correctly: 9 Documented Failure Modes
12 min readTypeSafe documents nine ways jev-1.13 gets things wrong. Here is what each one looks like in production and how to engineer around it.
The Traps: Noul vs Choice, Score Expectations, and Thresholds That Don't Transfer
11 min readJev's answers are calibrated, not algebraic. The invariants you assume between question types are not guaranteed, and TypeSafe publishes the numbers to prove it.
Four Architectural Patterns for Building with Jev
11 min readSpeculative fan-out, confidence-gated routing, composite scoring, and intent routing, with working code for each.
Atomic Questions: The Skill That Decides Whether Jev Works for You
10 min readDecomposition is the whole job. Broad questions hide judgments; atomic ones expose them so your code can inspect, tune, and combine them.
When Jev Beats an LLM, and When It Definitely Doesn't
10 min readA field guide to the tasks Jev is genuinely good at, the ones it cannot do at all, and the category error behind most of the viral demos.
Jev vs. Zero-Shot Classifiers, BERT, and Constrained Decoding
9 min readTaking the "this is nothing new" critique seriously in both directions, and working out what Jev actually replaces.