Skip to content

Jev Model Router — satviksinha

Claude Code routing mod that asks Jev which model tier and effort each turn needs, then directs that turn to haiku, sonnet, opus, or fable with visible route lines.

Stars
6
Language
TypeScript
License
MIT
Published by
Community
Author
satviksinha
Last verified
Oct 2026

About this project

Picks the model for each turn with Jev, TypeSafe's decision model. Supports both TypeSafe's direct API and the Vercel AI Gateway. You type a prompt. Before the turn runs, Jev is asked two questions at once: which tier should answer this, and how hard should it think. Every model request in that turn then goes to the model Jev named, and a line above the reply says which one.

Quoted from the project's README. Read the full README.

Highlights

  • Asks Jev tier plus effort before the turn runs and routes every model request in that turn to the named model.
  • Defines four tiers from trivial lookups to planning and debugging, with policy strings as the routing contract.
  • Prints route, confidence, timing, answering model, and cache share in replies, footers, and the /jev command.
  • Supports a quiet mode that hides markers without turning routing off.

Watch out

Fails open: a missing key, unpaid gateway account, or disabled hooks flag looks like routing doing nothing, so run the bundled check command. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

Other listings of this project

Forks and copies of this project in the directory. This entry is the canonical one.

2GitHub stars
Web chat where Jev picks the LLM for each message: one Choice question over four priced tiers, then the chosen model replies through the same Vercel AI Gateway key.
Jev Agents & MCP#community#awesomejev#catalog-import
2GitHub stars
A Python library that routes each prompt to the cheapest OpenRouter model that can handle it: Jev classifies domain and complexity, then a deterministic scoring function ranks OpenRouter's live catalog against that classification.
Jev Agents & MCP#community#awesomejev#catalog-import
2GitHub stars
Plugin for Claude Code, the Claude desktop app and Codex that has Jev read the task, then ranks live OpenRouter models by benchmark quality, price, speed and reasoning effort, returning a pick plus cheaper and faster alternatives with the exact command to switch.
Jev Agents & MCP#community#awesomejev#catalog-import

More Jev resources like this

20GitHub stars
TypeScript SDK and CLI coding agent where Jev only picks options while your code runs effects, building programs as syntax trees.
Jev Agents & MCP#community#awesomejev#catalog-import
13GitHub stars
Pi extension that classifies each task with Jev over OpenRouter and routes it to a Pareto-optimal model balancing measured quality, cost, and latency under a role policy.
Jev Agents & MCP#community#jevbest#catalog-import
12GitHub stars
Local OpenAI-compatible proxy that skips repeat model calls on same-intent paraphrases and retried agent turns, judged by Jev or a local System One model.
Jev Agents & MCP#community#awesomejev#catalog-import
10GitHub stars
Node router that sends each query to the cheapest capable LLM tier by asking Jev to classify demand against your tier descriptions, with no LLM call spent on routing.
Jev Agents & MCP#community#jevbest#catalog-import
7GitHub stars
Claude Code plugin that scores pending tool calls and finished answers with Jev noul probabilities, denying the call with the correction attached when a rule threshold is met.
Jev Agents & MCP#community#awesomejev#catalog-import
6GitHub stars
Terminal coding agent on Pi where Jev shortlists skills, ranks retrieved excerpts, and triages tool failures while the coding model handles reasoning and edits.
Jev Agents & MCP#community#awesomejev#catalog-import
Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai