[Evaluating agent responses with Jev, which does not

@LangChainJP LangChainJP · Sep 21, 2026 · case #591898 · listed via Jev Hub

Still from the clip posted with this case; watch it in the original post on X.

LangChain tested Jev as an evaluator for AI agent responses, using its typed answers and probabilities instead of generated text.

The original post, translated

[Evaluating agent responses with Jev, which does not generate text] LangChain, which provides an AI agent development platform, has published the results of an experiment using TypeSafe AI's model Jev as an evaluator. Jev does not generate text; it returns answers such as choices or numerical values and probabilities from the given information.

Show the original post in its source language ▾

【文章を生成しないJevでエージェントの応答を評価】 AIエージェント開発基盤を提供するLangChainが、TypeSafe AIのモデルJevを評価者に使う実験結果を公開した。Jevは文章を生成せず、与えられた情報から選択肢や数値などの答えと確率を返す。

The post above is a machine translation from ja; the untranslated text is in the fold-out.

Engagement when collected

Likes1

Bookmarks0

Reposts0

Replies0

Quotes0

Numbers are a snapshot taken from X when the case was added to the library (schema v1, collected 2026-09-19); they will not match today.

Where this case fits

Filed under agents & workflows. In the pattern Jev is built for, the model answers a bounded question per step — and ordinary code acts on the answer, because the answer is already a value rather than a paragraph. Other posts in the same family are on the agents & workflows page.

🤖 Agents & workflows