Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

jev-java

Idiomatic Java 17+ SDK for Jev with builders for structured instructions, rubric-bounded Choice options and Score levels.

Librarybeta
View on GitHub ↗

Overview

Jev Java SDK (olti1947) is a Java 17+ client library for TypeSafe's Jev, published under io.github.olti1947 version 2.0.0 for Maven and Gradle. A client created with the API key evaluates a state against a list of Choice, Noul and Score questions and returns typed answers, for example classifying a message's intent as cancel at 0.99 and checking whether the customer is angry with a helper that applies a threshold.

Wherever a primitive takes a string it also accepts structured values, following TypeSafe's advanced primitives docs: structured instructions with fields and extracted values, Choice candidates with what, not-for and example boundaries via a Rubric type, and Score levels with summaries and signals ordered low to high.

What the decision model decides

TypeQuestionAnswers
ChoiceSelect classificationbilling · support · cancel
Yes / noIs the customer expressing anger?—

How it works

  1. Add the Maven or Gradle dependency
  2. Create JevClient with the API key
  3. Call evaluate with state and typed questions
  4. Read typed answers and apply thresholds

Strengths

  • Idiomatic Java API
  • Supports structured instructions and rubrics
  • Published to Maven coordinates

Caveats

  • Community SDK

Getting started

Add `implementation 'io.github.olti1947:jev-java:2.0.0'`, then `JevClient.create(System.getenv("TYPESAFE_API_KEY"))`.

At a glance

Type
Library
Status
beta
Language
Java
Runtime
Java 17+ library
Has code
yes

Use cases and integrations

How curated lists describe it

Frequently asked questions

Which Java version?

Java 17 or newer.

Does it support rubric-style options?

Yes, via Rubric with what, not-for and examples.

Is it on Maven?

The README gives io.github.olti1947:jev-java:2.0.0 coordinates.

Related projects

  • ai

    Hack Club's lightweight LLM proxy, listed for adding a Jev forwarding endpoint behind its auth, limits and usage logging.

    MCP serverproductionTypeScript
  • Ax

    Ax, a DSPy-style multi-language LLM framework, listed with a native TypeSafe provider for Boolean, Choice, Score and raw Jev questions.

    PluginproductionTypeScript
  • Effect

    Effect TypeScript library whose @effect/ai-typesafe package maps classify, probability and rating operations to Jev questions.

    ProjectproductionTypeScript
  • instructor-php

    PHP monorepo whose Polyglot module adds a TypeSafe driver for typed Noul, Choice and Score decisions alongside LLM inference.

    ProjectproductionPHP
  • jev-mcp (freepik-company)

    Go MCP server giving any client typed Jev decisions via OpenRouter or TypeSafe, with validated answers and no paid retries.

    MCP serverproductionGo
  • llm-typesafe

    Simon Willison's LLM plugin that adds Jev as a model for yes/no, choice and score questions from the command line.

    PluginproductionPython