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

AgentScope

Alibaba's AgentScope 2.0 agent framework, which added a JevClassifierModel for binary, choice and score decisions via the TypeSafe SDK.

ProjectproductionApache-2.0
View on GitHub ↗

Overview

AgentScope 2.0 is a production-oriented agent framework from the agentscope-ai team that leans on models' reasoning and tool use rather than rigid orchestration. Recent releases added realtime voice, SOPs, a TeamPipeline for leader-member delegation, model routing middleware, A2A support and more.

In September 2026 it added Jev classifier support through the TypeSafe SDK via JevClassifierModel. The list summary describes this as native Jev classification for binary, choice and score routing across agent pipelines, so a pipeline can make typed decisions rather than prompting a chat model. The same release window added ModelRouterMiddleware for choosing a chat model per reply, TeamPipeline delegation from a leader agent to members, SOP support and A2A agents.

What the decision model decides

TypeQuestionAnswers
ChoiceWhich branch or agent should handle this step of the pipeline?(inferred)—
Yes / noBinary routing check within a pipeline(inferred)yes · no

How it works

  1. Configure JevClassifierModel with a TypeSafe key
  2. Use it inside a pipeline or routing step
  3. Receive typed probabilities
  4. Route the pipeline accordingly

Strengths

  • First-class integration in a widely used framework
  • Typed decisions alongside chat models

Caveats

  • Jev is one small integration among many features
  • Requires TypeSafe access

Getting started

See docs.agentscope.io and use JevClassifierModel via the TypeSafe SDK.

At a glance

Type
Project
Status
production
Language
Python
Runtime
agent framework
Has code
yes
Has tests
yes
License
Apache-2.0

Use cases and integrations

How curated lists describe it

  • Multi-agent platforms: multi-agent platform by Alibaba implementing native TypeSafe Jev classification models for binary, choice, and score routing across agent pipelines.

    yibie/awesome-jev

Frequently asked questions

How do I use Jev in AgentScope?

Through JevClassifierModel, which wraps the TypeSafe SDK.

What question types are supported?

Binary, choice and score, per the listing.

When was it added?

The news section lists it in September 2026.

Related projects

  • AgentConnect

    Open-source team agent platform for Slack, GitHub and more that asks Jev once per session which agent runtime and model to use.

    ServicebetaTypeScript
  • tiershift

    Policy-driven router that sends each LLM request to the cheapest capable model tier, escalating when Jev finds a reason.

    ProjectbetaTypeScript, Python
  • langgraph-jev-router

    LangGraph decorator that lets Jev take routing decisions it is confident about and falls back to your existing LLM router otherwise.

    LibrarybetaPython
  • Atomic

    Verifiable coding-agent runtime built on checkable execution graphs, listed with an optional Jev provider for bounded choices like routing.

    ServiceproductionTypeScript
  • LangChain

    LangChain's Python framework, which includes an optional TypeSafeClassifier runnable plus model-routing and risky-tool middleware using Jev.

    PluginproductionPython
  • LangChain.js

    LangChain.js, which includes an optional TypeSafeClassifier runnable and middleware for bounded routing and tool-call checks with Jev.

    PluginproductionTypeScript