An agent with a built-in decision layer
JevAgent: a System 1 + System 2 runtime
JevAgent pairs your agent with Jev, a decision model that makes quick judgment calls during a run: whether a request is clear enough to start, which specialist should take it, and whether the work is really finished. JevAgent is in preview, and its capabilities are rolling out one at a time.
Give generation and judgment different jobs
JevAgent pairs two models. Your agent reads the request, writes the reply, and uses its tools. Jev makes quick, structured judgment calls along the way and never writes the reply itself.
Keeping these jobs apart means the important decisions around a run, such as whether to ask the user a question first or whether the work is truly done, are made deliberately by a model built for judgment instead of being left to prompt instructions alone.
Two roles, one runtime
Current scaffold · PR #437System 1 · Generation
The agent does the work
Reads context, writes the reply, and uses its configured tools.
System 2 · Jev
A separate decision model
Designed to return structured judgments; it does not write the reply.
JevAgent runtime seam
One runtime boundary; the roles keep distinct responsibilities.
What you can use today
You can build on JevAgent in the SDK today. Jev's capabilities are rolling out one at a time: some are available, some are in preview, and some are still planned.
Create a JevAgent
You set up a JevAgent with two groups of settings. Creating an agent never calls Jev or costs anything; Jev only runs when your agent does.
Agent settings — Your agent's name, instructions, model, and tools, plus any specialist agents it can hand work to.
Jev settings — Optional. Choose which Jev capabilities to turn on.
A TypeSafe API key — Your agent can use any supported model provider. Jev runs on TypeSafe's decision model, so the capabilities that use Jev need a TypeSafe API key.
Create an agent
1from vidbyte import JevAgent, JevAgentSettings, JevPreflightPreset, JevRuntimeSettings
2
3agent = JevAgent(
4 JevAgentSettings(
5 name="support-agent",
6 system_prompt="Help the user resolve account questions.",
7 provider="openai",
8 model_name="gpt-4.1-mini",
9 ),
10 JevRuntimeSettings(
11 preflight=(JevPreflightPreset.CLARITY,),
12 ),
13)What Jev adds to a run
Turn on only the capabilities you need. Each one has its own page with setup steps, examples, and its current status.
Preflight questions
Self-alignment
Specialist routing
Continuation
Skill preloading
Usage and pricing
What a JevAgent run costs
Your agent's model is billed at its normal rate. Jev costs $0.042 per million input tokens, and its output is free.
Your own TypeSafe key — TypeSafe bills you directly.
Vidbyte's managed option — Jev usage is billed to your Vidbyte wallet at TypeSafe's rate plus a small markup.