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

ziru-jev-codex-router

Preview router for macOS Codex where Jev picks the GPT model and reasoning effort for each request through an isolated profile.

Projectbeta
View on GitHub ↗

Overview

ZIRU Jev Codex Router lets Codex choose the model and thinking effort for each stage of work. You select it with an explicit codex -p jev profile and work as usual; in Auto mode, Jev chooses between GPT-6 Luna, GPT-6.1 Sol and GPT-6 Astra along with how deeply the model should reason, and replies can show the chosen route. Other modes keep a manually chosen effort or a fixed model.

The aim is to use enough reasoning for each task, so a small edit and a final security review do not get the same treatment. The installer isolates the provider: normal Codex and the Desktop model picker keep their official configuration, and the router is only used through the explicit profile. An optional, build-specific Desktop integration adds four routing modes. It is a public prerelease (v0.1.0-preview.3) with documented validation and known limits.

What the decision model decides

TypeQuestionAnswers
ChoiceWhich model should handle this request?GPT-6 Luna · GPT-6.1 Sol · GPT-6 Astra
ChoiceWhat reasoning effort should be used?—

How it works

  1. Install and run codex -p jev.
  2. Each prompt goes through the local router.
  3. Jev picks model and effort (in Auto mode).
  4. Codex runs with the chosen route, optionally displayed.

Strengths

  • Per-request model and effort selection.
  • Isolated profile leaves normal Codex untouched.
  • Documented validation and limits.

Caveats

  • Prerelease preview.
  • macOS Codex Desktop integration is build-specific.

At a glance

Type
Project
Status
beta
Runtime
local provider/gateway for Codex (macOS)
Has code
yes

Use cases and integrations

How curated lists describe it

Frequently asked questions

What does Jev decide?

Which GPT model and how much reasoning effort to use per request.

Does it change normal Codex?

No, it is only used via the explicit codex -p jev profile.

Is it stable?

It is a public prerelease preview.

Related projects

  • jev-router★ 593

    Per-turn model routing for Claude Code and Codex: wraps the real CLI and lets Jev send simple turns to fast models and hard ones to strong.

    ProxybetaJavaScript
  • jev-codex-router★ 274

    Per-call router for Codex where Jev picks the model and thinking effort for every model call, including tool continuations.

    ProjectbetaPython
  • coding-router-jev

    TypeScript/Bun launcher that asks Jev which model and reasoning effort each Codex, Claude Code or Pi turn should use.

    ProjectbetaTypeScript
  • frost

    Go CLI model router that uses one Jev analysis to recommend a model, harness and effort level, then selects deterministically.

    CLIbetaGo
  • jev-codex-bridge

    Local bridge where Jev selects model and reasoning effort for Codex Desktop and CLI, with a Windows service and cache-aware routing.

    CLIbetaTypeScript
  • jev-codex-model-and-effort-router

    macOS wrapper for Codex that lets Jev pick the model and reasoning effort for a thread, then pins that route.

    Projectbeta