Skip to content
JevDirectory.org
Primitives

Criteria

The possible answers for a question: a map of options for Choice, an ordered list of levels for Score, or true/false for Noul.

Open docs.typesafe.aiAll 117 terms
Category
Primitives
Also known as
—
Related terms
5
Directory entries
2
Docs
docs.typesafe.ai
Added
2026-09-24

Definition

Criteria define the answer space. For a Choice they are a map of option to description; for a Score an ordered array of 2 to 10 level descriptions; for a Noul an optional description of what yes and no mean. Because the model returns a distribution over exactly these values, criteria are also the schema your code can trust.

Criteria are an extension of the instruction. Aligning the two matters: when instructions and criteria ask for different things, accuracy drops, and a Noul whose true maps to no performs worse than one phrased naturally.

Tagsprimitivesdesign

Related terms

Definitions that connect to this one.

InstructionsThe field that states what a question should judge; accepts a string, an object, or an array for extra structure.PrimitivesChoice optionOne key in a Choice criteria map, with a description that can be a string, a structured object, or null.PrimitivesScore levelOne entry in a Score's ordered criteria list; levels are numbered by array position from 0 and judged independently.PrimitivesNoul criteriaOptional true and false descriptions that define what a yes and a no mean for a Noul question.PrimitivesContradictory criteriaWhen instructions and criteria ask for different things, accuracy drops; treat criteria as an extension of the instruction.Evaluation & Training

From the directory

Primitives: Choice, Score & Noul
The three TypeSafe question types, the typed answers they return, how to choose between them, and how to ask several in a single call.
Practices & PatternsDocs#official#primitives#choice
Official
API Reference
The raw HTTP contract behind every SDK: POST a state plus typed noul, choice, and score questions to /v1/systemone, and get one answer per question, with error codes and retry guidance.
Sites & GuidesDocs#official#docs#api
Official

From the community

Posts from builders shipping with Jev right now.

Follow @typesafeai

An open-source BS meter for debates and investor calls

Chetaslua
Chetaslua
@chetaslua
·

🚨 Open Source Jev BS meter you can use this to analyze any debate / investor call / interview / sales pitch / podcast video fact check live , for example this dario interview cost 60 Jev calls / 111K tokens / $0.0047 github.com/ChetasLua/jevm…

Watch on X
Chetaslua
Chetaslua
@chetaslua

🚨 I gave the Trump vs Kamala debate a live BS meter using Jev every sentence, both candidates, 5 yes/no questions each 1,191 Jev calls / 1.18M tokens / 415 ms median total cost : $0.0497 same questions for both, clips picked by one fixed rule, not a fact-check

Watch on X
Reply
Read 15 replies

A Jev-shaped model on Cerebras and Qwen

Shannon 🤖☠️
Shannon 🤖☠️
@iamMrDuncan
·

Built an alternative version of @typesafeai but on @cerebras with Qwen 3.8 27b. Similar quality, similar performance, but vastly different cost. TypeSafe was way cheaper, and did beat Qwen on performance. Closest we can get using LLMs I think. Source: github.com/iammrduncan/ty…

Watch on X
Diogo Almeida
Diogo Almeida
TypeSafe AI
@CompleteSkeptic

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x

Watch on X
Reply
Read 15 replies

Jev benchmarked on two public safety corpora

Gaurav Gosain
Gaurav Gosain
@_GauravGosain
·

1/ Benchmarked TypeSafe's Jev on two public safety corpora. It doesn't generate text, it returns calibrated probabilities you threshold in code. 96.5% on prompt injection, all 662 messages in deepset/prompt-injections. No tuning. 325ms p50.

Terminal dashboard showing 96.5% accuracy on prompt injection and 89.0% pairwise on vulnerable code, with a context ablation and calibration plot.
Diogo Almeida
Diogo Almeida
TypeSafe AI
@CompleteSkeptic

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x

Watch on X
Reply
Read 1 reply

An open 151M-parameter decision engine

Hemant
Hemant
@heman10x
·

TypeSafe AI came out of stealth with Jev, and access is behind a waitlist. I built an open source version Verdict (Open-jev) you can run right now in a browser tab: And its a real post trained model..(link in comments) It is a post trained 151M parameters model. ModernBERT-base Show more

Image
Diogo Almeida
Diogo Almeida
TypeSafe AI
@CompleteSkeptic

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x

Watch on X
Reply
Read 2 replies

Agent Handoff Gate: agents verify what they hand off

A local Telegram analyzer with Jev as the classifier

Dagmawi Babi
Dagmawi Babi
@DagmawiBabi
·

I vibe-coded this simple, open-source and local Telegram content analyzer using Jev as a classifier. Jev Classifier • jevclassifier.vercel.app Export your channel data as JSON and it will analyze each post's intent, quality, sentiment, and reaction tone and if it's a DM/Group Show more

Watch on X
Diogo Almeida
Diogo Almeida
TypeSafe AI
@CompleteSkeptic

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x

Watch on X
Reply
Read 2 replies