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

jev-llm-router

Python CLI that shows which OpenAI model Jev would route a query to, independently reviewed by a local Ollama model.

Projectprototype
View on GitHub ↗

Overview

Jev LLM router is a small Python project for studying how Jev chooses an OpenAI model for a text query. Jev selects a route, a local Ollama model independently reviews the same query, and the CLI shows both choices, probabilities and review status. It evaluates the routing decision only and does not call the selected OpenAI model, so no OpenAI key is needed.

It requires Python 3.10+, a TypeSafe key and Ollama running locally with a text model such as llama3.1:8b. Setup instructions cover macOS, Linux and Windows, and environment variables must be set in each shell since .env is not loaded automatically.

What the decision model decides

TypeQuestionAnswers
ChoiceWhich OpenAI model should handle this query?—

How it works

  1. Enter a text query
  2. Jev chooses a route with probabilities
  3. A local Ollama model reviews the same query
  4. CLI shows both decisions and agreement

Strengths

  • No OpenAI key required
  • Independent local cross-check

Caveats

  • Study tool; does not execute routes
  • Ollama model quality affects the review

Getting started

`python -m pip install -e .`, set TYPESAFE_API_KEY and OLLAMA_MODEL, then run the CLI.

At a glance

Type
Project
Status
prototype
Language
Python
Runtime
CLI with local Ollama
Has code
yes

Use cases and integrations

How curated lists describe it

Frequently asked questions

What does Jev decide?

Which OpenAI model a query should go to.

Does it call OpenAI?

No, it only displays the selected model.

What is Ollama used for?

An independent local review of the same query.

Related projects

  • instinct-ai

    Open-source dual-brain runtime that answers typed decisions with a fast local engine or the Jev API before falling back to LLMs.

    ServicebetaPython, TypeScript
  • jev-laya-dsh

    Local Laya decision layer for DeepSeek Harness: routes prompts, gates risky tool calls and swaps models, all offline at no token cost.

    Projectbeta
  • jev-switch

    Unofficial Rust gateway routing /v1/systemone requests to TypeSafe, Vercel, OpenRouter or local Laya via an editable DAG.

    ProxybetaRust
  • Keel

    Local-first macOS coding workspace that runs your agents and lets a local Laya or hosted Jev selector pick a route the host validates.

    ServicebetaRust
  • one-system

    Go gateway that puts local and hosted System One models behind one TypeSafe API, with a local Laya router choosing Laya or Jev.

    ServicebetaGo
  • s1-auto-router

    Claude Code plugin where a local Decider model on your Mac picks the model and effort level for each subagent task.

    PluginbetaPython