Routing
Not every task needs your most expensive model.
Let JEV help your agent choose the right route, thinking depth, or speed for the job.
Your models. Your priorities. Your rules.
JEV MCP
Meet JEV, TypeSafe’s System One decision model, accessible through Viorant.
Give it context. Ask the question. Get a choice, a score, or a yes/no with a probability your agent can act on.
No extra prose. No need to change how you build. Just connect JEV through Viorant and keep working with the tools you already use.
Agents make hundreds of small decisions per run. Which model handles this step. Is this ticket urgent. Did the cheap answer hold up. Is this page safe to read. Most teams answer each one by asking a chat model to write a sentence and then parsing the sentence. It costs output tokens, it takes seconds, and the answer comes back as prose with no probability attached.
JEV cuts through the noise. A structured decision. A probability to work with. And the freedom to set your own threshold.
The intelligence is there. The final call stays yours.
Not every task needs your most expensive model.
Let JEV help your agent choose the right route, thinking depth, or speed for the job.
Your models. Your priorities. Your rules.
Start small. Escalate when it matters.
Let a smaller model take the first shot. JEV evaluates the result against your context, helping your agent decide when a bigger model deserves the next turn.
You set the bar. Your workflow makes the call.
Score a support ticket for urgency, frustration or risk against your own rubric. Escalate above the line, auto-handle below it.
Screen a fetched page or a tool result for prompt injection before the agent reads it. Block an action whose probability of being correct falls under your bar.
Don't let every claim go unchecked.
Give JEV the draft and the evidence behind it. Get a structured assessment of what holds up, what conflicts, and what needs another look.
Your evidence. Your standards.
Three primitives and three tools built on them. Every call returns a probability.
jev_noulA batch of propositions in, a probability per proposition out.
jev_choiceOne multiple-choice question you wrote, the chosen option with a probability per option.
jev_scoreOne question against an ordered rubric, the expected score with a probability per level.
jev_verifyClaims plus an evidence text; each claim comes back verified, contradicted or unsupported with a confidence.
jev_screenRaw external text; returns prompt-injection and substance flags and, if you state the task, relevance.
jev_findRanks candidates against a plain-language query. No embeddings, no index.
Jev generates no text, so there are no output tokens. You are billed on what you send.
A coding agent gets a turn. It calls jev_choice with the task and the model list. Jev answers "small" at 0.91. The turn runs on the cheap model. The expensive one only sees turns Jev was unsure about.
A support bot drafts a reply with a small model. It calls jev_verify with the reply's claims and the ticket thread as evidence. One claim comes back contradicted. The draft goes to the large model for a rewrite. The other nine replies that hour never did.
A research agent fetches a third-party page and passes the raw text to jev_screen with the task purpose. The screen flags an instruction hidden in an HTML comment and rates the page low on relevance. The agent skips it without reading the instruction.
Your JEV. Your workflow. No starting over.
Already building with JEV? Bring it into the tools you know through Viorant’s hosted MCP connection.
One URL. One Viorant account. One credit balance for JEV and other eligible Viorant services.
Keep your setup. Keep your momentum. Build on your terms.
Jev is a standard remote MCP server with OAuth. If your client connects to one, it connects to Jev. The agent does not have to run on Viorant.
Jev sits next to every other connector your agents use: browse the Viorant connector catalog.
Jev is a standard remote MCP server with OAuth. Each client below gets its own config. To deploy an agent that calls Jev, install the Viorant CLI.
requires:
- connector: https://jev.viorant.io/mcp
scopes: []
auth: oauth
kind: connector
tools:
- jev_verify
- jev_screen
- jev_find
- jev_noul
- jev_choice
- jev_scoreDeclare Jev in agent.vio’s requires block, deploy, then run vio connect once to sign in. The agent calls Jev’s tools mid-run like any other connector.
vio docs agents
vio deploy .
vio connect <deployment-id>vio docs agents prints the Jev recipe. Paste its requires block into agent.vio, deploy, and sign in with your Viorant account when vio connect opens the browser.
https://jev.viorant.io/mcpIn the Hub, open Connectors, choose Add connector, paste this URL on the MCP URL tab, name it Jev, and sign in with your Viorant account.
claude mcp add --transport http jev https://jev.viorant.io/mcpThen run /mcp and choose Authenticate for jev.
{"mcpServers":{"jev":{"url":"https://jev.viorant.io/mcp"}}}Add this to mcp.json, or wire it from Cursor’s MCP settings panel.
{"servers":{"jev":{"url":"https://jev.viorant.io/mcp"}}}Add this to your MCP settings.
{"mcpServers":{"jev":{"serverUrl":"https://jev.viorant.io/mcp"}}}Add this to your MCP settings.
codex mcp add jev --url https://jev.viorant.io/mcpThen run codex mcp login jev to sign in.
{"mcpServers":{"jev":{"httpUrl":"https://jev.viorant.io/mcp"}}}Add this to settings.json’s mcpServers.
https://jev.viorant.io/mcpAny client that supports a remote HTTP MCP server with OAuth connects the same way: point it at this URL and sign in with your Viorant account when prompted.
Sign in with your existing Viorant account. There is no separate Jev account.
More JEV. More room to build.
Viorant’s Free Plan comes with 2,500 one-time Vio Credits. Use them to explore JEV through our hosted MCP connection, without paying for a subscription.
At 100 Vio Credits per million JEV input tokens, that’s up to 25 million* input tokens in theory.
No text output tokens to pay for. No need to change your existing MCP workflow.
And your credits work across other eligible Viorant services, too.
*Actual JEV token capacity may be lower due to rounding or other credit usage. Free Plan and Program Terms apply.
Intelligence, On Your Terms.
New to Viorant? Start with the Free Plan, connect JEV through MCP, and put your included credits to work.
Your Free Plan credits are already included. No separate promotional credit grant is required.
Compare Viorant credits and plans.
No. Your Viorant account signs in to Jev over OAuth.
No. Any client that supports a remote HTTP MCP server with OAuth connects the same way.
Simple. JEV uses your Vio Credits at 100 credits per million input tokens at the launch rate. No text output-token charges.
Viorant’s Free Plan already includes 2,500 one-time Vio Credits. That’s theoretically enough for up to 25 million JEV input tokens if you use them exclusively for JEV.
Actual capacity depends on rounding and how you use your credits. You can top up when you need more.
What you send is what counts: your questions, context, choices, and any evidence JEV needs to evaluate.
A short ticket uses fewer input tokens than a long document. You decide what context to send, and your Vio Credits cover the usage.
Technical limits and billing details are available in the documentation.
No. It returns a decision and a probability, never prose. Use it where your code needs a number to act on.
Your tools. Your workflow. Your JEV.
Ready to put JEV to work? Pick your MCP client, follow the guide, and connect through Viorant.
JEV is developed by TypeSafe and accessible through Viorant’s hosted MCP integration.
JEV results and performance are not guaranteed. Usage is subject to Viorant’s Terms of Service and applicable pricing and service conditions.
The Viorant Builder Credits Program promotes JEV usage through the standard Free Plan credits. No additional promotional credit allocation is included.